GENDER DIFFERENCE AND ITS IMPACT ON AGRICULTURAL

PRODUCTIVITY: THE CASE OF WENCHI DISTRICT IN SOUTH

WEST SHOA ZONE, ETHIOPIA

THESIS SUBMITTED TO

ALEMAYA UNIVERSITY

SCHOOL OF GRADUATE STUDIES

IN PARTIAL FULFILMENT OF THE REQUIREMENT FOR

MASTERS OF SCIENCE DEGREE

IN AGRICULTURAL ECONOMICS

BY

WAKWEYA TAMIRU YADA

June 2004

Alemaya University
 ALEMAYA UNIVERSITY

SCHOOL OF GRADUATE STUDIES

GENDER DIFFERENCE AND ITS IMPACT ON AGRICULTURAL

PRODUCTIVITY: THE CASE OF WENCHI DISTRICT IN SOUTH WEST SHOA

ZONE, ETHIOPIA

BY

WAKWEYA TAMIRU YADA (B.Sc.)

Approval by Board of Examiners

______________________________ ______________________
Chairman Signature

Bezabih Emana (Ph.D.) ________________________

Major Advisor Signature

Milkesa Wakjira (Ph. D) ________________________

Co-Advisor Signature

_________________________________ _________________________
External Examiner Signature

_________________________________ _________________________
Internal Examiner Signature
 BIOGRAPHIC SKETCH

The author was born in June 1977, in Illubabor zone of Oromiya Regional State, Bure

district. He attended his education in Gore Elementary and Secondary School and

completed his high school studies in 1995.

In September 1995, he joined Alemaya University (the then Alemaya University of

Agriculture) and graduated with B.Sc. degree in Agricultural Economics in July 1999.

Immediately upon successful completion of university education, the author was

employed by World Vision International/Ethiopia as pha se in program design team and

served for one year at Wenchi district, south west Shoa. And then, he was transferred to

Tigray Regional Sate in the same organization and served as grant and relief projects

officer for two years.

In 2001/2002 academic year, he joined the School of Graduate Studies at Alemaya

University in the Department of Agricultural Economics as a self sponsored student.

ii
 ACKNOWLEDGEMENTS

My greatest thank and heartfelt appreciation goes to Dr. Bezabih Emana, major thesis

research advisor, for his whole hearted and unreserved comments, advice and stimulating

guidance. He has been devoting much of his time from his busy schedule for giving me

consultation. I also greatly acknowledge his assistance and provision of related study

materials starting from the very inception of the proposal, questionnaire design up to the

completion of the thesis. I am also deeply indebted to Dr. Milkesa Wakjira, co-advisor,

for his valuable and constructive comments, which apparently enabled me to produce this

paper in its present shape and content. Without their assistance, the completion of this

paper would have been hardily possible.

As self-supporting student, I owe much to my family. I am also very grateful to Ato

Shimelis Abate, WVE Relief program Manager, for encouraging me to attend the

Graduate Studies while I was on job. Last but not least, I would like to thank staff of

WVE, who in many ways encouraged me to pursue the study.

iii
 TABLE OF CONTENTS
Page
BIOGRAPHIC SKETCH --------------------------------------------------------------------------------------- ii

ACKNOWLEDGEMENTS ------------------------------------------------------------------------------------ iii

TABLE OF CONTENTS --------------------------------------------------------------------------------------- iv

LIST OF TABLES ------------------------------------------------------------------------------------------------ vi

LIST OF FIGURES ----------------------------------------------------------------------------------------------vii

LIST OF APPENDICES ----------------------------------------------------------------------------------------vii

ACRONYMS AND ABBREVIATIONS ------------------------------------------------------------------viii

ABSTRACT ---------------------------------------------------------------------------------------------------------- x

CHAPTER ONE: INTRODUCTION ------------------------------------------------------------------------ 1

1.1 Background ---------------------------------------------------------------------------------------- 1

1.2 Statement of the Problem---------------------------------------------------------------------- 4
1.3 Objectives of the Study ------------------------------------------------------------------------- 6
1.4 Significance of the Study ----------------------------------------------------------------------- 6
1.5 Scope and Limitations of the Study --------------------------------------------------------- 7
1.6 Organization of the Thesis --------------------------------------------------------------------- 8
2.1 Conceptual Framework ------------------------------------------------------------------------- 9
2.1.1 Gender and its Roles ------------------------------------------------------------- 9
2.1.2 Gender Gap/Disparity ---------------------------------------------------------- 11
2.1.3 Household Headship ------------------------------------------------------------ 11
2.2 Role of Gender in Agriculture ---------------------------------------------------------------15
2.3 Measurement of Productivity in Agriculture --------------------------------------------18
2.3.1 Productivity--------------------------------------------------------------------- 18
2.3.2 Production Functions ----------------------------------------------------------- 21
2.4 Review of Empirical Findings----------------------------------------------------------------26
CHAPTER THREE: AN OVERVIEW OF THE STUDY AREA ----------------------------------31

3.1 Location --------------------------------------------------------------------------------------------31

3.2 Physical Features and Area Coverage -----------------------------------------------------31
3.3 Population -----------------------------------------------------------------------------------------32
3.4 Economic Activities -----------------------------------------------------------------------------32

iv
 3.5 Infrastructure and Marketing Services ---------------------------------------------------33
CHAPTER FOUR: METHODOLOGY--------------------------------------------------------------------36

4.1 Source and Type of Data ----------------------------------------------------------------------36

4.2 Sampling Technique and Sample Size -----------------------------------------------------36
4.3 Methods of Data Analysis ---------------------------------------------------------------------37
4.3.1 Cobb-Douglas Production Function-------------------------------------------- 37
4.3.2 Estimation Technique and Testing Procedures --------------------------------- 40
CHAPTER FIVE: RESULTS AND DISCUSSIONS ---------------------------------------------------43

5.1 Descriptive Analysis ----------------------------------------------------------------------------43

5.1.1 Household Characteristics------------------------------------------------------ 43
5.1.2 Farm Characteristics ----------------------------------------------------------- 48
5.1.2.1 Land Use Pattern----------------------------------------------------------- 48
5.1.2.2 Livestock Holding ---------------------------------------------------------- 51
5.1.2.3 Labour Utilization ---------------------------------------------------------- 52
5.1.2.4 Use of Fertilizer, Herbicides and Improved Seed -------------------------- 54
5.1.3 Crop Yield----------------------------------------------------------------------- 55
5.1.4 Gender Based Division of Labour ---------------------------------------------- 56
5.1.5 Sources of Farm Income-------------------------------------------------------- 60
5.1.6 Access to and Control Over Productive Resources----------------------------- 61
5.1.6.1 Land ------------------------------------------------------------------------ 61
5.1.6.2 Oxen ------------------------------------------------------------------------ 64
5.1.6.3 Institutional Services ------------------------------------------------------- 65
5.1.7 Decision Making in the Household --------------------------------------------- 68
5.2 Econometric Analysis --------------------------------------------------------------------------70
5.2.1 Definition of Variables and Hypothesis Setting -------------------------------- 70
5.2.2 Estimation of the Production Function ----------------------------------------- 72
5.2.3 Comparison of Productivity of the two Groups--------------------------------- 77
5.2.4 Marginal Value Product of Inputs---------------------------------------------- 82
5.2.5 Source of Productivity Difference ---------------------------------------------- 84
CHAPTER SIX: SUMMARY AND POLICY IMPLICATIONS -----------------------------------88

6.1 Summary-------------------------------------------------------------------------------------------88
6.2 Policy Implications ------------------------------------------------------------------------------90
REFERENCES ----------------------------------------------------------------------------------------------------92

APPENDICES -----------------------------------------------------------------------------------------------------98

v
 LIST OF TABLES
Page
Table 1. Summary of Estimates of Male-Female Differences in Technical Efficiency.... 29

Table 2. Distribution of Sample Households by PA and Sex of Household Head............ 36

Table 3. Average Family Size of the Households in AE................................................... 44

Table 4. Average Family Size by Age Group and Sex of Household Head (AE)............. 45

Table 5. Literacy Status of Household Head (%) .............................................................. 46

Table 6. Marital Status of the Sample Households (%).................................................... 47

Table 7. Reasons for Being Female Headed (dejure) ........................................................ 47

Table 8. Land Use Pattern of the Households (Mean in ha).............................................. 49

Table 9. Cropping Pattern of Sample Farmers .................................................................. 50

Table 10. Livestock Ownership by Gender (TLU)............................................................ 52

Table 11. Sources of Labour Use in Man-days per ha ...................................................... 53

Table 12. Strategies Used to Overcome Labour Shortage (%).......................................... 54

Table 13. Mean of Quantity Applied and Cost of Agricultural Inputs Applied ................ 55

Table 14.Yield of Major Crops by Gender (Qt/ha) ........................................................... 55

Table 15. Proportion of Male and Female Family Member Participated (activity profile)

.................................................................................................................................... 57

Table 16. Source of Cash Income by Gender .................................................................... 60

Table 17. Proportion of Households Acquiring Land From Different Sources (%) ......... 63

Table 18. Type of Land Access by Gender ....................................................................... 64

Table 19. Access and Control Profile of Resources and Benefits ..................................... 68

Table 20. Proportion of Households Participating in Decision Making (%)..................... 69

Table 21. Multicollinearity Test Among Explanatory Variables ...................................... 73

vi
Table 22. Parameters of Cobb-Douglas Production Function........................................... 74

Table 23. Comparison of the Overall Production Parameters of Both Groups ................. 78

Table 24. Estimates of Production Function for Pooled Sample With Dummy Gender ... 79

Table 25. Chow’s Test for Checking the Difference in the Slope Parameters .................. 80

Table 26. Estimates of a Pooled Production Function With Both Intercept and Slope

Dummies..................................................................................................................... 81

Table 27. MVP and Factor Cost of Significant Explanatory Variables ............................ 82

Table 28. Geometric Mean of Dependent and Exp lanatory Variables in the Model ........ 85

Table 29. Decomposition of Productivity Difference between MHH and FHH ............... 85

LIST OF FIGURES

Figure 1.Technical Efficiency and Allocative Efficiency of Male and Female Farmers--21

Figure 2. Map of Oromiya Regional State by Zone--------------------------------------------34

Figure 3. Map of South West Shoa by District-------------------------------------------------35

Figure 4. Distribution of Land Size (ha)---------------------------------------------------------49

LIST OF APPENDICES

Appendix 1. Survey Questionnaire---------------------------------------------------------------94

Appendix 2. Conversion of Livestock into TLU----------------------------------------------102

Appendix 3. Conversion Factor for Family Member into AE ------------------------------102

Appendix 4. Conversion Factor for household Labour into ME----------------------------102

vii
 ACRONYMS AND ABBREVIATIONS

ADP Area Development Program

AE Adult Equivalent

CD Cobb-Douglas

CISP International Committee for the Development of Peoples

CSA Central Statistical Authority

DA Development Agent

FAO Food and Agriculture Organization

FHH Female Headed Household

GDP Gross Domestic Product

Ha Hectare

Kg Kilogram

Km2 Kilo Meter Square

Km Kilometre

ME Man Equivalent

MEDaC Ministry of Economic Development and Cooperation

MHH Male Headed Household

MoFED Ministry of Finance and Economic Development

MVP Marginal Value Product

NOP National Office for Population

PAs Peasant Associations

PCICRW Population Council and International Center for Research on Women

Qt Quintal

viii
SNNP South Nation and Nationalities People

SPSS Statistical Package for Social Science

TFP Total Factor Productivity

TLU Tropical Livestock Unit

VIF Variance Inflation Factor

WB World Bank

WVE World Vision International Ethiopia

ix
 GENDER DIFFRENCE AND ITS IMPACT ON AGRICULTURAL

PRODUCTIVITY: THE CASE OF WENCHI DISTRICT IN SOUTH WEST SHOA

ZONE, ETHIOPIA

By: Wakweya Tamiru Yada (B.Sc.)

Advisors: Bezabih Emana (Ph.D)
Milkesa Wakjira (Ph.D)

ABSTRACT
The study ex amined the gender difference and its impact on agricultural productivity in
Wenchi district of south west Shoa zone, located at 125 km south west of Addis Ababa,
between Ambo and Welliso towns. The specific objectives were to assess the extent to
which the agricultural production system is gender oriented; to examine access and
control over productive resources; and to estimate men’s and women’s productivity in
agriculture.

Cross-sectional data collected from a total of 140 respondents were used in this study
whereby 65 were female headed and 75 were male headed households. The data were
analyzed using Statistical Package for Social Science (SPSS). Descriptive statistics such
as frequency, mean, percentage, t-test and chi-square were used to summarize and
compare the information between the two groups. Moreover, Cobb-Douglas(CD)
production function was used to estimate the productivity difference in agriculture
between male and female headed households.

Results of the study showed that male headed households (MHH) own more of productive
resources such as land, livestock, labour and other agricultural inputs as compared to
female headed households (FHH). Moreover, the estimate of CD production function
shows that livestock, herbicide use, land size and male labour were statistically
significant for MHH while livestock, land size, herbicides use and female labour were
significant variables for FHH.

The comparison of the marginal value product (MVP) with the factor cost showed that
MHH could increase productivity using more herbicides and male labour while FHH
could do so by using more herbicides, male and female labour. The agricultural
productivity difference between MHH and FHH was about 68.83% in the study area.
However, if FHH had equal access to the inputs as MHH, gross value of the output would
be higher by 23.58% for FHH. This may suggest that FHH would be more productive
than MHH if they had equal access to inputs as MHH.

Based on the results obtained, the following policy implication can be drawn: accessing
FHH to inputs that increase the productivity of agriculture such as herbicides, livestock
and male labour; increasing the productivity of land; and introducing technologies that
reduce the time and energy of women especially for enset processing.

x
CHAPTER ONE: INTRODUCTION

1.1 Background

Agriculture is the major economic sector and the main source of livelihood for the majority of

the people in Ethiopia. No other country in sub-Saharan Africa and only two other countries in

the world, derive a higher share of gross domestic product (GDP) from agriculture than Ethiopia

(Block, 1999). According to CSA (2002), 85 percent of the total population of the country lives

in rural areas and 90 percent of the population depends on the agricultural sector. Furthermore,

the sector contributes 50 percent to the GDP, 70 percent to the export earnings and employees 80

percent of the workforce. Hence, the overall performance of the Ethiopian economy is highly

influenced by the performance of the agricultural sector. All in all, economic development in this

country cannot be envisaged without the development of agriculture and hence poverty reduction

is closely linked to the development of this sector (MoFED, 2002).

Despite the importance of agricultural sector in Ethiopia, its performance has remained below its

potential. Information from MEDaC (1998) indicates that the average growth rate of agriculture

during the years 1974 to 1992 was merely 2 percent, which was lower than the rate of population

growth. Similarly, the average rate of growth of this sector was 2.27 percent per annum between

1992 and 1998 (NBE, 1999). In both periods, the rate of agricultural growth is much lower than

the population growth, which is 3 percent annually. This low growth rate resulted in a

considerable food deficit in the country. Hence, Ethiopia imports food substantially.

1
Rapid population growth, small size and fragmentation of land, poor infrastructure, traditional

agricultural practices and implements, extension services which are inadequately funded,

shortage of and poor distribution of agricultural inputs, gender related and other personal

characteristics of the producers (Seyoum et al., 1998) and degradation of land compounded by

weather related factors such as drought or rain failure (MoFED, 2002) have contributed to the

poor performance of agricultural sector in Ethiopia. In general, these problems can be classified

into technological, environmental and policy problems (SG, 1999).

Given this low productivity, a significant number of people have suffered from food shortage.

Obviously, women and children are the usual victims of the problem. On the other hand, women

are the majority of the world’s producers. They make up to 60 to 80 percent of the agricultural

workers in Africa and Asia (Davidson, 1995) and more than 40 percent in Latin America (John

and David, 1991). Most African women are farmers and responsible for 70 percent of staple

food production (World Bank, 1989 as cited by Assefa and Mutambirwa, 1992).

Like in most developing countries, women in the rural parts of Ethiopia account for 50 percent of

rural population. About 88 percent of the country’s women live in rural areas, nearly 85 percent

of their labour spent on agricultural activities such as food processing, storage, weeding,

harvesting, marketing of produce, preparing trashing field and caring for animals (Bogalech,

2000). However, in spite of their substantial contributions to the development of the economy in

general and agricultural sector in particular, they are constrained by a number of factors.

Therefore, it is important to pay special attention to the work and lives of the poor rural women.

If any development strategies aimed at increasing production fail to address women’s role and

2
needs, it is unlikely to succeed (Koopman, 1997). In other words, achievement of substantial

growth in production depends on understanding of the intra-household separation and interaction

of men’s and women’s activities (Martha, 1987).

Within a given ecological environment, agricultural productivity is determined by the amount of

land, capital and other inputs used and by the quality of these factors. As a general proposition,

given technologies and managerial skills are the same, farmers who have identical access to

similar factors, both in quantity and quality, will produce identical outputs of a given crop. That

is, their productivity will be identical. However, men and women rural households especially in

developing countries have different access to technologies, factors of production and support

services. If they use different technologies or different quantities of these factors or there are

differences in the quality of these factors, their productivity will differ (Saito et al., 1994).

Overcoming agricultural stagnation and food insecurity mainly depends on increasing

agricultural productivity. In Ethiopia, like in many other parts of sub-Saharan Africa, where

subsistence agriculture predominates, placing strong emphasis on increasing the productivity of

labour, land, capital and other resources is of paramount importance. Therefore, this study is

initiated in view of understanding the agricultural productivity difference between men and

women farmers in Wenchi district of south west Shoa zone of Oromiya Re gional State.

3
1.2 Statement of the Problem

About 75 percent of the world’s population lives in the developing countries and also 75 percent

of the population of developing countries lives in rural areas. Out of these, more than 50 percent

are women (World Bank, 2003). Similarly, women in Ethiopia constitute about half of the total

population (CSA, 2002).

According to World Bank (2000a), about 1.2 billion people in the developing countries live in

conditions of absolute poverty, that is, in a situation where the per capita income is short of

meeting enough food to provide 2250 calories per person a day. More than half of the poor are

women. Especially, female headed households (dejure) are at risk of living in poverty and

disproportionately represented among the poorest of the poor, the uneducated and the first to

suffer from drought and famine (Duggan, 1997; Quisumbing, 1993; World Bank, 2000b ; Saito

and Surpling, 1992).

In general, female headed households have less access to land, labour, well paying jobs and

government services including credit, educational opportunities (Ellen, 1992; World Bank,

2000b). In most African countries female headed households are the first to suffer when

economic, political and environmental deterioration occur and the last to gain when there are

improvements (Elabour, 1991). For example, in some countries, including Kenya and Malawi,

the poorer the household, the more likely it is to be headed by a woman (World Bank, 1989i as

cited by Uma, 1991; Buvinic and Gupta, 1996).

4
Moreover, the proportion of women and their children who live in poverty has been increasing.

Reasons for the increase are multifaceted, including the break up of the traditional family system

and the rise of single headed households; the sex specific migration resulting in left behind

female heads in place of origin and the transformation of the global economy, which has shifted

patterns of employment (Lisa and Ritu, 1995; Buvinic and Gupta, 1996). Since the number of

female headed households in developing countries is growing, it posses significant challenges in

designing policies to improve household food security and especially child nutrition (Ellen,

1992).

The same phenomena occur in Ethiopia. Female headed households have limited access to and

control over resources, which is crucial to effectively and efficiently meeting their strategic and

practical needs due to social and cultural biases in the society (Tiruwork, 1998; Addis, 2000;

Dejene, 1994). They are also not homogenous. There are differences in terms of access to and

control over resources and decision-making between these female headed households.

In general, women’s contributions in rural Ethiopia have remained invisible; especially female

headed households are more invisible to researchers, donors and policy makers (Addis, 2000;

Tiruwork, 1998). These situations have put women at a disadvantageous position with respect to

agricultural resources, leading to low productivity of female headed households. This will further

decrease their participation in economic activities. Therefore, it is necessary to find out the

difference in the productivity of male and female headed households.

5
The situation in the study area, Wenchi district, is not different. The district is one of the highly

populated areas in south west zone where population density is about 214 persons per Km2 ,

whereas it is about 137 persons per Km2 for the zone (CSA, 2002). As the result, farm size is so

small to produce sufficient food to the population and the number of female headed households

is also increasing from time to time since males are migrating to other areas in search of better

employment leaving behind their wives and children. As the result, females will takeover the

position of their husband in addition to their routine household management. Therefore, it is

essential to study the productivity of female headed in agricultural sector vis-à-vis their male

counterparts in the area where a gender dis-aggregated information in this aspect is missing.

1.3 Objectives of the Study

The overall objective of the study was to analyse differences in productivity of male and female

headed households in agricultural production in Wenchi district of south west Shoa zone of

Oromiya Regional State. The specific objectives were:

1. to assess the extent to which the agricultural production system is gender oriented,

2. to examine the differences in access to and control over productive resources between male

and female headed households, and

3. to estimate male and female productivity in agriculture.

1.4 Significance of the Study

As agriculture is the core of the country’s economy, it can contribute much to the development

through the improvement of the productivity of the sector. Understanding the difference between

6
the househo lds especially male and female headed households helps to design means of

developing the agricultural sector and thus essential for the long-term success of the economy.

On the other hand, the causes of gender related differences in productivity is crucial, because, if

gender affects the productivity directly, meaning, if men’s and women’s productivities are

different when they have exactly the same constraints and resources, then it may be necessary to

modify research strategies to ensure increasing their productivity. If, on the other hand,

differences in productivity arise because men and women face different constraints, especially

access to complementary inputs, then it may be more important to design projects that improve

women’s access to these complementary inputs. Therefore, the result of this study enables us to

know the sources of agricultural productivity differences between men and women farmers

working in a similar environment.

The study, furthermore, provides some basic information needed by policy makers and

institutions interested in designing programs and projects that are appropriate to the needs of

both men and women. Development actors operating in the area may also benefit from the result

of the research so that they work to fill the gap existing between men and women towards

bringing sustainable development in the area.

1.5 Scope and Limitations of the Study

To compare agricultural productivity differences between men and women, it may be more

appropriate to disaggregate the data by gender of the plot manager, rather than by gender of

household head (Saito, et al., 1994; Udry, 1994). However, when compared to other African

7
countries, Ethiopia appears to be unique in having relatively strong unitary households because

even though men and women perform different agricultural activities they generally cultivate the

same plots (CISP, 1997). In the study area, too, women and men accomplish farming activities

on the same plots. There is no separately identified plot managed by male and female in the farm

households. Therefore, comparison of productivity difference in agriculture was made at

household level rather than at plot level.

Despite appropriateness of using survey method of data collection, the other limitation of this

study was that the respondents were asked to quantify the value of continuous variables, such as

the yield of major crops, the size of land he/she owned, etc from their memory. The actual values

for the continuous variables would have been more appropriate to take measurement of the

variables. Moreover, respondents were informed about the purpose of the study before the

interview conducted.

1.6 Organization of the Thesis

The contents of this thesis are classified into six major chapters. The first chapter consists of

background to the research, research problems, the objectives, the significance as well as the

scope of the study. The second chapter deals with the review of literatures on topics relevant to

the study. The third chapter presents an overview of the study area. The fourth chapter deals with

the research methodology. Results, both descriptive and econometric, are presented and

discussed in the fifth chapter. The last chapter summarizes the major findings of the study and

draws appropriate policy implications.

8
CHAPTER TWO: REVIEW OF LITERATURE

2.1 Conceptual Framework

Under this section an attempt is made to discuss some of the concepts used in this study such as

gender, gender role, gender gap and household headship.

2.1.1 Gender and its Roles

Gender is one of the ways in which societies and smaller social groups are stratified. It is a term

used to describe social differentiation (Laura, 1991). In other word, gender is a established social

and cultural role of men and women or it is a system of roles and relationships between men and

women that are determined by the social, political and economic concepts. It is the analytical

term, which refers to socially determined difference between women and men as opposed to sex,

which is biologically determined.

According to Kabbar (1997) cited by Bogalech (2000a) gender is that people are born female or

male but learn to be girls and boys who grow into women and men. They are taught what the

appropriate behaviour and attitudes; roles and activities are for them and how they should relate

to other people. This learned behaviour makes up gender identity and determines gender roles.

Gender role stems from the socio-cultural construction of what a female or a male is expected to

do, perform or take responsibilities in a given cultural context. It refers to determined patterns of

behaviour in terms of rights, duties, obligations and responsibilities assigned to female and male

in a given society.

Roles differ in different settings and culture of a society. It is also interdependent, learned,

dynamic (changes over time) and multi- faceted (differ within and between cultures). Since

9
gender is created by society, its meaning varies from society to society and changes overtime

(John and David, 1991). Many authors argue that cultural ideology about men’s and women’s

roles is a critical factor in determining the way gender relations of production are ordered in a

given society and, therefore, must be seriously considered in development planning (Tiruwork,

1998).

In all societies, men and women play different roles, have different needs and face different

constraints. Gender roles differ from the biological roles of men and women, although they may

overlap in nearly all societies. Women’s biological roles in child bearing may extend their

gender roles to child rearing, food preparation and household maintenance. In agriculture,

women are actively involved in production in most countries; however, men’s and women’s

roles differ widely across regions. Among some groups, for example, women are responsible for

milking; in others, men do this work.

Gender roles demarcate responsibilities between men and women in social and economic

activities, access to resources and decision- making authority. Biological roles are fixed but

gender roles can and do shift with socio-economic and technological changes. For example, the

introduction of new crops and technologies, mounting pressure on land, or increasing poverty or

migration can change the roles of men and women in agriculture. Social factors can reinforce or

decrease gender-based disparities (World Bank, 2001).

Gender roles can be categorized into three (Ellis, 1988). These are productive, reproductive and

community roles. Productive role involves the production of goods and services for consump tion

and trade. Reproductive role include bearing and rearing children and taking primary

10
responsibility for domestic maintenance. Communal roles involve the collective organization of

social events and services. So, the analysis in this paper is related mostly to the productive role

of gender.

2.1.2 Gender Gap/Disparity

There are socio-economic indicators of gender inequality. These include measure of

employment, education, health, ownership of property and income disparities. Gender gap results

from inequality in decision making power which leads to inequality in access to resources and by

the differential treatment given to women and girls as compared to that given to men and boys.

Gender discrimination exists as part of the social system and runs through all aspects of life and

at different levels such as at family level, community level and institutional level (Bogalech,

2000b).

Gender inequality imposes costs on productivity, efficiency and economic progress. By

hindering the accumulation of human capital in the home and labour market and by

systematically excluding women or men from access to resources, public services or productive

activities, gender discrimination diminishes economy’s capacity to grow and to raise living

standard (World Bank, 2001). Accordingly, the study attempts to analyse the gender

discrimination regarding access to productive resources and decision- making power in the study

area.

2.1.3 Household Headship

The problems of defining household and head of household are well known (Buvinic and Gupta,

1996). Household mean different things to different people in different places. There is also a

growing debate on the desirability of generating definitions, which may be universally

11
applicable. However, certain definitions are given in some data reports, census and studies. The

Central Statistical Authority (CSA, 2002) defines the term household as an aggregate of persons,

generally but not necessarily bound by ties of kinship, who live together under the same roof and

eat together or share in common the household food. A household is composed of a head,

relatives living with him/her and other persons who share the common life for reasons of work or

other considerations. They may be related or unrelated persons, or a combination of them.

Furthermore, a person who lives alone is considered as a separate household, as also a person

who makes arrangements for his/her own food.

There is a wide variation of the definition of the term household head. According to some

literatures (e.g. Chant, 1997; Rosenhouse, 1988; Buvinic and Youssef, 1978) there are two

categories of the concept of head of household, namely breadwinning (chief earner) and decision

making (power/authority). Proponents of the “chief earner” concept defined household head in a

way it helps to identify the groups of households in need of government assistance programs.

However, chief earner may not have a significant status in the family. Although women could be

playing the major economic role in the household, patriarchal traditions and the practice of self

or proxy reporting may mean that men may be reported as household head irrespective of their

earnings (Buvinic and Youssef, 1978).

On the other hand, supporters of the “power/authority” definition link the concept with a broader

topic of gender relations within the household and believe that the term reflects the power

hierarchy of the household more than just who earns more. In this view, it is the control of the

household resources and day-to-day engagement that is more important, not who earns more.

Evidences from many studies have suggested that “people can make major economic

12
contribution to households without even being present in the household” (Mencher and

Okongwu, 1993). In general, the debate about the content of the concept of household head is

centered around the question whether the household head is the person who has the highest

authority and decision- making power, or whether he/she has control over the general affairs of

the family unit including decision-making concerning its economic, social and political affairs

than other members of the household (Medina et al., 1994).

Female headed households include households headed by single mother, divorced, separated or

widowed women. In general, these can be categorized as dejure and defacto headed households

(Grown and Sebstad, 1989; Ellen, 1992). In the literature, the distinction between defacto and

dejure headship is common. Defacto female headed households are those households where the

male heads are absent for more than 50 percent of the time. In these households, husbands or

other male relatives often play a role in basic decision-making and make varying contributions to

household income. In dejure female headed households, women are legal and customary heads

of the households. They are usually headed by single women, those who are divorced, separated

or widowed. Typically, in dejure households, the heads of the households are likely to have

control over most household income and asset (Ellen, 1992).

In recent years, the number of female headed households has been increasing rapidly throughout

the world, in countries with very different socio-economic backgrounds and cultural traditions.

The incidence of female headed households has reached a degree not known before in many

countries in north America, in Latin American countries, in Africa and Europe as well as in Asia

(Buvinic and Youssef, 1978; Folbre, 1991; Buvinic and Gupta, 1996). In the United States,

Canada and Northwestern Europe about 20 percent of all households are believed to be headed

13
by women (Folbre, 1991). In the third world, it is estimated that approximately one third of all

households are headed by women, even though spatial variation is marked (Buvinic and Youssef,

1978; Rosenhouse, 1988). Central America and sub-Saharan Africa are regions with the highest

proportions of female headed households. In Africa, the percentage of female headed households

is around one-third of the total households and in Zambia it approaches to one-half (Uma, 1991).

In countries of south east Asia, there is evidence of a marked increase in female supported, if not

female headed households, that are beginning to be harder hit by economic crises and economic

adjustment policies (PCICRW, 1998).

A considerable proportion of female headed households and its increasing trend are common

features of most countries, the patterns of female headship are very different in different regions

and even in different countries of the same region. In the United States, Canada and many other

developed countries of northwestern Europe, the incidence of female headed households is

accounted mainly for the increase in number of divorces, separations, single-mother households

and widowhood (Garfinkel and Mclanaha, 1986; Folbre, 1991). In Africa, a large number of

female headed households concentrate in rural areas since the male heads migrate to urban areas.

In Latin America, it is women who migrate to cities causing the number of female headed

households in urban areas high (PCICRW, 1998). In south Asia (India, Bangladesh, Pakistan,

etc.) widowers and left behind wives account for most of female headed households (Mencher

and Okongwu, 1993).

In general, at any point in time, one third of the world’s households are female headed, either

temporary due to a male partner’s migration or permanently due to separation, divorce or death

(Duggan, 1997). Rates of female headship in today’s less developed countries are much higher

14
than those that prevailed in presently industrial countries (Oestergard, 1992 as cited in Duggan,

1997). Similarly, in Ethiopia the number of female headed households are increasing from time

to time. For example, in 1994 it was about 22.2% (CSA, 1998) and increased to 25.3% in 2001

(CSA, 2001).

In this study, the definition of households given by the Central Statistics Authority (CSA, 2002)

is adopted. Moreover, in this study a woman is categorized as female headed household when

she is registered as head of a household by peasant association, pay land use tax in her name or

has a separate dwelling unit without a male partner.

2.2 Role of Gender in Agriculture

All over Africa men and women have separate responsibilities and play different but

complementary roles. Though the gender division of labour differs considerably across border

depending on culture and economic status, women universally carry the major burden of

producing food and providing food daily for consumption to the family (Davidson, 1995).

Boserup (1970) argued that the number of women taking part in agric ultural work was found to

be higher than that of men. The reasons are, firstly, older men do often stop working by leaving

it to their usually younger wives or to their children; secondly, many old women are widows who

must feed themselves; thirdly, more young boys than girls go to school from the villages. She

also showed that women’s agricultural production was critical in sustaining local and national

economies.

15
Women in all circumstances of class and marital status- single, married, divorced or widows-

participate in different agricultural activities. They play an important role in the performance of

agriculture. However, much of the work that women do is invisible or is either assumed to have

of little value. This is because most tasks and even occupations done primarily by women take

place within the household and are not seen as productive (Visvanathan, 1997).

There is no question that African women provide most of the labour required to produce the food

consumed in Africa with much regional variations (Uma, 1991). Women contribute an estimated

three quarter of the labour required to produce the food consumed in Africa, mostly under

conditions of primitive, hand hoe technology and low labour productivity. Women produce more

than half of the locally grown crops in developing countries and as much as 80 percent of it in

Africa (ibid).

As the result, women in developing countries work longer hours than men (John and David,

1991; Uma, 1991; Saito et al., 1994). In Africa, 15 of the 17 studies summarized by Brown and

Haddad (1995) cited by Ilahi (2000) found out that women work more than men do. Similarly,

in sub- Saharan Africa rural women work, on average, fifty percent more hours per day than men

(Saito et al., 1994). In Tanzania, rural women work for 12-16 hours a day all year round while

men sometimes do not even manage half this level of labour input (Tibaijuka, 1994). Women in

the third world now carry a double or even triple burden of work as household work, childcare

and subsistence food production, in addition to the expanding involvement in paid employment

(John and David, 1991).

16
Furthermore, aggregate data indicate that African women provide about 90 percent of the labour

for processing food crops and providing household water and fuel wood; 80 percent of the work

in food storage and transport from farm to village; 90 percent of the work in hoeing and

weeding; 60 percent of the work in harvesting and marketing, 60 percent of domestic animal

care, 50 percent of planting, 30 percent of ploughing (Davidson, 1995; FAO, 1985 as cited by

Uma, 1991). In general, African women typically work up to 16 hours per day due to their

diverse and numerous responsibilities (ibid). It is also estimated that 78% of the women in Africa

are active in agriculture compared with only 64% of men (Gittinger et al., 1990).

Similarly in Ethiopia, women play an important role in the production of food at household level

(Dejene, 1989; Zenebework, 2000). With the exception of ploughing and sowing, women

participate in every aspects of farm production. They also play a significant role in animal

husbandry in all rural areas and especially in pastoral areas of the country (Zenebework, 2000).

Rural women are engaged in the agricultural and domestic activities on average for 15 to 18

hours a day in Ethiopia (NOP, 1999).

According to the study made by Wudinesh (2003) in Amhara (North Wollo and Humera), Tigray

(Eastern and Southern), SNNP (North Omo), female farmers provide more than half (50-58%) of

the total labour force and time inputs required for crop production in the surveyed areas.

Similarly, they cover up to 77% of the total labour force and time inputs required in livestock

production. Likewise, female farmers are not only involved in crop and livestock production but

also in generating additional income for their families. They produce goods such as storage

17
containers, baskets, etc. They are also engaged in brewing and in petty trading of agricultural

produces.

Furthermore, female farmers play key roles in maintaining the daily life of their families and

maintaining the social cohesion of families. About 33% and 67% of the overall labour force and

time inputs required for household work, farm production, income generation and off- farm

activities, etc are covered by husbands and wives, respectively. Therefore, as indicated in

different studies, Ethiopian women farmers are the key actors in maintaining the daily lives of

their families and in contributing to the rural economic development as a whole (Wudinesh,

2003). So, this study identifies the major areas of participation of both men and women in the

agricultural production system and also identifies gender-oriented activities in the study area.

2.3 Measurement of Productivity in Agriculture

2.3.1 Productivity
In economic theory, productivity is defined in terms of the rate of output produced per unit of

input utilized, if the production process involves a single input and output. Beyond the single

output and single input case, however, the definition and measurement of productivity become

less straightforward. In this case, the average productivity concepts of comparing an aggregate

output index to an aggregate input index is used to obtain a ratio measure of productivity (Antle

and Capalbo, 1988; Coelli, 1998).

The birth of modern theory of productivity took place in 1957, when Robert Solow started the

profession by the findings of his research, which won him noble prize (Solow, 1957).

Productivity is the major component of growth. The importance of productivity in economic

18
development is universally recognized. The economic achievement of some of the developed

countries is attributed to increase in productivity. Changes in productivity become more

important for countries where the resources are limited in supply and have very high social

opportunity cost. According to Cheema (1978), productivity can be achieved through change in

efficiency and technological progress.

According to Wong (1986), the story of productivity that the ratio of output to inputs is at the

heart of the record of men’s efforts to raise himself from poverty. He reported that the dynamic

forces in agricultural sector are both the causes and consequences of agricultural productivity

changes. These dynamic forces include technological progress, accumulation of physical capital,

improvement of human capital and innovation of institutional arrangements. Each of these forces

has a different role to play in the process of agricultural development and productivity growth.

The primary use of productivity as stated by Wong (1986) is to measure the effectiveness with

which producers utilize their resources for the purpose of production (as cited by Asmare, 1989).

In general, productivity can be measured in two ways: Productivity in relation to a particular

input or partial productivity and productivity in relation to all inputs together or total factor

productivity (TFP). Labour and capital productivity, which may be measured by the value added

per unit of labour and per unit of capital, respectively, are partial productivities. The partial

productivity method enables us to measure the contribution of each input, which could be hidden

in the total productivity measure.

19
There are two distinct approaches to measurement of productivity: the growth accounting and

econometric methods. The growth accounting approach involves compiling detailed accounts of

inputs and outputs, aggregating them into input and output indexes and using these indexes to

calculate a TFP index. One of the major problems of computing a TFP index is aggregating the

different inputs and outputs. In most cases, outputs can be aggregated in terms of their monetary

values; however, the problem lies in the aggregation of inputs. This approach also does not allow

for identification of the contribution of individual factor of production. In addition, success of the

use of TFP method depends very much on the availability of data of all components of the index

(Coelli, 1998).

Alternatively, econometric methods can be used to estimate the compone nts of TFP using

production, cost and profit functions. There are various production functions that enable

measurement of productivity. Both the growth accounting and the econometric approaches of

measuring productivity have strengths and limitations. Each approach also requires certain

assumptions, which must be considered in interpreting the findings of productivity studies (Antle

and Capalbo, 1988). In econometric approach, the assumptions depend on the type of production

function used. Furthermore, in order to estimate cost or profit functions, the additional

assumptions of competitive pricing and efficient utilization must be made. The growth

accounting approach in addition requires price information on both inputs and outputs. However,

both approaches can be useful and should be considered appropriate. Therefore, the choice

between the two approaches should be based on research objectives, the data requirement, the

availability of data and the appropriateness of assumptions (ibid).

20
2.3.2 Production Functions

The relationship between inputs and outputs in a production process of a firm or any production

unit can be investigated through the application of a production function. According to Heady

and Dillon (1961), a production function is a technical relationship between output and factors of

production of any production unit. In other words, a production function shows the maximum

output that can be produced from given quantities of input with a given state of technology. The

production function, therefore, represents the technology used by the farmers based on the

technical methods of production.

Such inputs and output relationship in a production function is often represented by the following

function:

Y = f ( X 1 , X 2 , X 3 ,......... ......, X n , U ) --------------------------------------------(1)

Where, Y denotes the output, Xi is the amount of the ith input factors, f is the functional form

relating to the output and the ‘n’ variable inputs and U is disturbance term.

Using production function it is possible to measure technical efficiency (TE) and allocative

efficiency. Given the technology and set of input levels, TE reflects the ability of a farmer to

produce output. Thus, it is associated with the farmers’ ability to equate the level of its actual

production from a given input to the maximum possible level of production that could be

produced using the same level of inputs. Suppose that male and female farmers have the same

input levels but male farmers are more technically efficient. For the same level of inputs, say X0 ,

the quantity produced by male farmers would be greater than that produced by female farmers

21
(Figure 1). This implies that male farmers can technically produce more output for each level of

inputs relative to female farmers. The female farmers' production function may be below the

male production function because they may use traditional technologies, due to lack of

knowledge, lack of access to modern inputs, or higher costs to adopting the new technologies.

Allocative efficiency, on the other hand, refers to the adjustment of inputs and outputs to reflect

relative prices. It measures the ability of farmer to equate the marginal value product (MVP) per

unit of input across different outputs. If the marginal cost (MC) of an input is equal to its

marginal revenue, the firm is said to be allocatively efficient and if its marginal cost is greater

than its marginal revenue, the firm is said to be allocatively inefficient. A farm with allocative

efficiency minimizes the total cost of producing a given quantity by selecting a combination of

factor inputs where the slope of the production function is equal to the slope of the cost function.

The difference between the two types of efficiency could be illustrated by using one factor

product space as follows.

Figure 1.Technical and Allocative Efficiency of Male and Female Farmers

Mc 2

A Mc1
Male (F2 )
D
B
Female (F1 )

X0 X1
Source: Ellis (1988)

22
Assume that a certain farm has a given amount of inputs to produce only one output, say Y. F1

and F2 are two different production functions, where F2 represents the frontier production

function and is greater than F1 . F2 displays higher output for all positive level of input use than

F1 . Therefore, F2 is technically superior to F1 . A farm producing Y3 at point D on F2 has higher

TE compared to point C, where the maximum level of production (Y3 ) is produced using the

same amount of factor input (X0 ). But comparison in terms of allocative efficiency shows that

producing at point C on production function F1 is allocatively inefficient compared to a farm

producing Y1 at point B on the same production function but the gained efficiency at increased

level of input (X1 ). The ratio Y0 /Y3 measures the TE of the farm producing at point C.

At point B, the farm produces at a point where the MVP of the factor input is equal to its MC,

thus the farm is said to be allocativelly efficient. The level of allocative efficiency of the farm

producing at C could be measured by using the output ratio, Y0 /Y1 . At point C the farm is both

technically and allocatively inefficient. At point B, the farm is allocatively efficient but

technically inefficient while at point D the farm is technically efficient while it is allocatively

inefficient. It is at point A that the farm is both technically and allocatively efficient. At this

point the farm is said to be economically efficient (Farell, 1957).

Generally, the improvement in male farmers’ efficiency over that of female farmers could be due

to technical change or the exploitation of scale economies or from some combination of these

factors (Coelli, 1998). Technical efficiency can be obtained by following the best techniques,

which involve the efficient use of inputs, given the technology. In other words, technical

23
efficiency is determined by the method of application of inputs, regardless of the levels of inputs

(Kalirajan et al., 1996).

Studies based on production function can guide resource allocation. For instance, within

production function, production elasticities of inputs can be evaluated in the elasticities of

substitution between inputs, and returns to scale can be examined. These characteristics can be

tools for economic analysis and policy formulation (Saito et al., 1994).

There are diversified forms of production functions applied in economic literature, all of which

have their own quality and drawbacks. The most common ones are linear, the Cobb-Douglas

(CD), the quadratic, the translog and the constant elasticity of substitution (CES) produc tion

functions. The choice of a function, however, involves a certain amount of subjective judgement;

guided by considerations of a prior economic and physical logic; goodness of fit; ease of analysis

and judgement about economic implications (Heady and Dillon, 1961). No one function is

superior in every aspect and it is difficult to show conclusively that one particular function is the

correct one (ibid). Considering relevance of the study, comparability with previous studies,

flexibility and computational ease as criteria for choosing the model, CD production function

was selected among the other functions.

CD production function is one of the most widely used functions in the economic analysis of

problems related to empirical productivity estimation in agriculture and industry. Many

empirical studies including Moock (1976), Bindlish and Evenson (1993), Saito et al. (1994),

Udry, et al. (1995) and Addis et al. (2000) have employed the CD form of production function to

measure agricultural productivity.

24
The CD production function may be good approximation for production process for which

factors are imperfect substitutes over the entire range of inputs value (Heady and Dillon, 1961).

It has also a number of desirable properties like the coefficients are positive and each less than

one. The sum of elasticities of output with respect to the relative inputs also provides the returns

to scale of the parameters. Although this function has other advantages in that it shows

diminishing marginal return, it involves some limitations. One of the limitations of the CD

production function is that the elasticity of substitution between factor inputs is restricted to

unity. The implications of zero output at zero inputs may also be unacceptable in some instances.

However, since the advantages out weigh the limitations, it is one of the commonly used

production functions.

The CD production function has the following features where some of them make it so

interesting and popular. First, the function is homogenous. The sum of the parameters has

interesting economic interpretation since it gives information about the returns to scale or the

scale of operation of the production process. The returns to scale are increasing, constant or

decreasing depending on whether the sum is greater than one, equal to one or less than one,

respectively. Secondly, the function is strictly quasi concave for positive values of inputs; and its

isoquants are negatively slopped throughout and strictly concave for positive value of inputs.

Thirdly, the function yields diminishing return to each input, i.e. the value of the production

elasticities are less than one. Finally, the coefficients are the output elasticity coefficient for

inputs and show the relative distributive shares of inputs in the total output. So, because of these

and others, CD production function is selected for this study.

25
2.4 Review of Empirical Findings

Structural or institutional factors may contribute to gender differences in productivity as

agricultural systems are modernized. Boserup (1970) made the classic feminist argument linking

resources to productivity. Although differences in productivity between the sexes might be

expected to fall as agriculture becomes less dependent on human muscular power, she observed

that men’s labour productivity tends to rise while women’s remains static. She concluded that the

tendency towards a widening productivity gap is often exacerbated by cash crop cultivation

among men, while women produce food crops for the family without cash income for investment

in farming technique.

Different studies have been conducted using different econometric models to reveal whether men

are more productive/efficient than women in farm production. Some of these, which have

relevance to this study, are briefly discussed here under.

A study of 152 maize farmers (51 were female farmers) in Kenya's Vihiga district in the 1970s

estimated yield functions for all farms with a female manager dummy, for male and female-

managed farms separately. The study found out that women, who make up a third of the sample,

are at least as productive as men: the female farmer dummy, though positive, was not significant

by itself in the pooled regression. Interactions between the female farmer dummy and other

inputs suggest that while women benefit less from more densely planted farms; they tend to

make better use of labour on maize farms than men do. Exposure to the extension service is

associated with greater technical efficiency if the farmer is male and not too well educated;

although there are efficiency gains for women with at least a primary education. Results from

26
separate regressions for male and female- managed farms indicate that primary education has a

positive and significant effect on yields for women but negative and significant effect for men

(Moock, 1976).

Bindlish and Evenson (1993) had undertaken studies in Kenya and Burkina Faso using a total of

675 households for Kenya (241 were female) and 2336 households for Burkina Faso (173 were

female). The results of these studies indicated that coefficient of gender dummy (1=female

headed household; 0=otherwise ) in the pooled data was insignificant indicating, in general, male

and female farmers are equally efficient as farm managers. In this study, the coefficient of

primary education was negative but insignificant.

Udry et al. (1995) examined gender based productivity differences in Burkina Faso using CD

production function. The data were drawn from a four-year panel study (1982-1985) of 150

households. They found that output per hectare was lower on plots controlled by women.

However, such differences in productivity cannot be taken as indicative of production

inefficiency. The authors noted that the gender yield differential, apparently, is caused by the

differences in the intensity of input use on plots controlled by men and women rather than by

differences in the efficiency with which these inputs are used. The estimates of the parameters

imply that, on average, households could increase output of crops (those crops grown by both

men and women) by about 10 percent, by reallocating labour and manure used on men's plots to

women's plots (Udry et al., 1995).

Another study from Kenya was based on a survey conducted in three districts (Kakamega,

Muranga and Kilifi) in 1989-90 using CD production functions with the gross value of maize,

27
beans and cowpeas per hectare as the dependent variable. The regression for all plots showed

that, while positive, the coefficient for male plot manager dummy was insignificant. The result of

this study also suggests that the gross value of output per ha was 8.4 percent higher for men than

for women, however, if women had the same human capital endowments and used the same

amount of inputs as men, the value of their output would have increased by 22 percent (Saito et

al., 1994).

Another study made in Nigeria (Oyo state) found out that the coefficient of dummy gender for

male headed household was negative and insignificant for the data taken from farm household

level, suggesting that there is no difference in technical efficiency between male and female

farmers but were significantly higher for men when total value of production at the plot level was

used (Saito et al., 1994).

Finally, Quisumbing (1995) reviewed seven studies that estimated differences in technical

efficiency between male and female farm mangers or household heads using production

functions and concluded, in general, that male and female farmers are equally efficient as farm

managers. However, their productivity is below potential. Capturing this potential productivity

gain by improving the circumstances of women farmers would substantially increase food

production in sub-Saharan Africa, there by significantly removing one determinant of food

insecurity in the region. Because wo men produce an estimated 75% of sub-Saharan Africa’s

food, if the results from Kenya were to hold in sub-Saharan Africa as a whole, simply raising the

productivity of women to the same level as men could increase total production by 10 to 15 %

(Saito and Surpling, 1992).

28
Table 1. Summary of Estimates of Male-Female Differences in Technical Efficiency

Authors/year Total No. of Gender variable Coefficient Dependent variable

sample female
Moock, 1976 152 51 Female 0.09 Log of maize output per
(Kenya) manager hectare
dummy=1
Bindlish and 675 241 Female head -0.022 Ln of value of total
Evenson, dummy=1 output at household level
1993 (Kenya)
601 182 Male farmer 0.126 Ln of value of crop
dummy=1 production at plot level
Saito et al., 453 147 Male headed -0.017 Ln of value of crop
1994 (Kenya) dummy=1 production at household
level
Bindlish and 2336 173 Female head -0.146 Ln of crop production at
Evenson, dummy=1 household level
1993 (Burkina
Faso)
225 15 Male headed -0.130 Ln of value of crop
dummy=1 production at household
Saito et al. level
1994(Nigeria) 1174 289 Male farmer 0.559* Ln of value of crop
dummy=1 production at plot level
Jamison and 1363 134 Male head 0.95**
Lau, 1982 Mechanical dummy=1
(Korea) farms Log of value of
541 Non 67 Male head 0.059 agricultural crop output
mechanical dummy=1
farms
Tiruwork, 500 250 Male headed 0.076
1998 (East dummy=1 Ln of quantity of wheat
(Ethiopia) Gojjam) and teff produced at
720(North 360 Male headed 0.267* household level
Shoa) dummy=1
Source: Own Summary (2003) NB: *, **=Significant at 1% and 5% probability level, respectively.

29
There are limited empirical studies on gender difference in agriculture in Ethiopia. Study

conducted by Tiruwork (1998) in north Shoa and east Gojjam zone using CD production

function concluded that female headed households are less productive than male headed because

of lack of productive resources such as land, adult male labour and extension service. Dummy

variable representing sex was one of the significant explanatory variables for north Shoa zone

while not for east Gojjam zone. The positive and significant coefficient for male headed dummy

sex in north Shoa indicated that male headed household are more technically efficient than that

of female headed households.

Another study conducted by Addis et al. (2000) on gender differentials in agricultural

productivity among smallholders in east Shoa using cross sectional data from 180 households in

1996/97 concluded that male headed households had more land, labour, capital particularly

livestock and access to formal education compared to female headed households. The gross

value of the output was 1.3% higher for the female headed households if the average values of

inputs from male headed households were used. This suggests that no significant difference

would exist if female headed households had equal access to inputs as male headed households.

According to Dejene (1994), differential access to productive resource is one of the factors
explaining iter-household variations in productivity. He concluded that female farmers, in
general, appear to be less productive than male farmers using descriptive statistics. However, he
did not consider if female farmers have equal access to productive resources as that of their male
counterparts. Therefore, this study will contribute to the limited information on agricultural
productivity differences between men and women farmers in Ethiopia in general and in the study
area in particular.

30
CHAPTER THREE: AN OVERVIEW OF THE STUDY AREA

3.1 Location

Wenchi district is one of the ten districts found in south west Shoa zone of Oromiya Regional

State (Figure 2). It is located at a distance of 125km south west of Addis Ababa, the capital of

the country, between Ambo and Welliso towns. The capital town of the zone is Welliso, which is

9 km from the capital of the district, Chitu. The district is surrounded by three districts; Ambo in

the north, Ameya in the west and Welliso in the south and in the east (Figure 3).

3.2 Physical Features and Area Coverage

Wenchi district is characterized by undulating hill with plain mountain topography ranging

between 1500-3387 meter above sea level (m.a.s.l). It mainly consists of dega (37% of the PAs

of the district), woinadega (46%) and kolla (17%) agro-climatic zone and the average rainfall

varies between 1200-1420 mm per year. Average temperature varies between 100 C and 300 C.

The district covers an area of 475.51 km2 (CSA, 2002) and divided into 23 PAs and one rural

town. According to information obtained from the agriculture office of the district, about 72.5%

of the land is cultivated, 13.7% grazing land, 9.7% natural forest, 1% water body and 3.1%

belongs to other categories.

The water resource found in the district comprises of rivers such as Walga, Borochu, Genetu,

Amagna, Delana, Dedebiya and many others scattered in the district. There is also one creator

lake, Wenchi Lake, in the district.

31
3.3 Population
Currently the total population of the district is 97,856 out of which females account for 50.8%.

The community in the district comprises of a total of 18,568 households (27% female headed

households). The average family size is estimated to be 5 persons per household and the average

population density is 214 per km2 making it the largest densely populated district in the zone

followed by Welliso (CSA, 2002).

The population of Wenchi district is almost exclusively Oromo constituting 97% of the

population. The rest 3% of the ethnic groups are Guraghe and Amhara. The religions of the

district are Orthodox Christian, Protestant and Muslim constituting 93%, 5% and 2% of the

population, respectively.

3.4 Economic Activities

Agriculture is the dominant economic activity engaging 90% of the labour force (CSA, 2002).

Crop production is dependent on rainfall and the major crops produced in the area according to

their importance are teff1 (26% of total area cultivated), wheat (24%), barely (20%), maize (6%),

sorghum (4%)and pulses (horse bean, field pea and lentil covering about 11% of the total

cultivated area). Enset 2 is the only perennial and staple food crop produced in all areas of the

district. Productivity of these crops is below the national average due to poor fertility of soil,

poor agronomic practices and use of low level of agricultural inputs.

Livestock are also reared by most families. Oxen provide traction power for the cultivation of the

agricultural lands. On the other hand, livestock are kept as a source of income through milk,

1
Teff- one of the indigenous cereals grown in Ethiopia
2
Enset- a banana like green fibrous food crop widely grown in south west and southern central Ethiopia.

32
butter, meat and egg production and as a means of savings. Livestock productivity is also low

due to absence of adequate feed and disease infestation.

With regard to the off- farm activities practiced in the area, small percent of the total population

is engaged in this sector of the economy. According to the base line survey of World Vision

Ethiopia conducted in 2001, around 21% of the population is engaged in off- farm activities. Pot

making, tannery, blacksmith, weaving, tailoring, carpentry and petty trading are some of the off-

farm activities widely practiced in the area. All craftsmen practice traditional technology

inherited from their parents that require a greater energy and more time resource. The level of

income generated by the participants of this sector is considerably low.

3.5 Infrastructure and Marketing Services

The existing social services found in the area include two health clinics, one health post, eight

elementary and junior schools (1-8 grade), four 1-7 grade schools, one 1-6 grade and one 1-4

grade school. There is one semi- manual solar telephone center in Chittu. The Welliso-Ambo and

Welliso-Amaya all weather roads pass through the district and provide major transport facilities

for the area.

There are three large local market places in the district: one at Werabu village near Dariyan

town, one at Lemen village and the other one is at Buri village. There are also 6 small and 6

medium markets. When farmers want to sell some of their farm produces, they have to travel

long distances to reach the market places. They use either pack animals, loading over their back

(women), shoulder (men) or incurring high transportation cost to use vehicles.

33
CHAPTER FOUR: METHODOLOGY

4.1 Source and Type of Data

Both primary and secondary data collected from Wenchi district were used. Primary data

were collected from sample households through structured questionnaire (Appendix 1) and

secondary data from concerned line offices such as agricultural office, education office,

Wenchi Area Development Program (ADP) and from administration office of the district.

The questionnaire covered information on household demographic and farm characteristics,

crop and livestock production, household income and ownership of farm inputs. Both male

headed and female headed households in the sample PAs were interviewed. A pre testing of

the questionnaire was conducted before actual data collection was made. The interview was

conducted by five enumerators who were trained on the subject matter of the questionnaire.

The data collection was carried out between October and November 2003.

4.2 Sampling Technique and Sample Size

A two stage random sampling technique was used to select the sample households in the

study area. The first stage was simple random sampling of 6 PAs from the 23 PAs found in

the district. Then from these PAs a total of 75 male headed and 65 female headed households

were randomly selected. Hence, a total of 140 households were interviewed (Table 2). The

size of the sample was limited to 140 because of shortage of logistics and time.

Table 2. Distribution of Sample Households by PA and Sex of Household Head

Name of PA Total Sample
Male headed Female headed Male headed Female headed
Meti Walga 652 158 13 11
Dimitu Godeti 417 151 8 10
Lemen Methora 610 170 12 12
Waldo Talfam 938 173 20 12
Fite Wato 695 172 14 12
Belbela Bulbula 372 103 8 8
Total 3684 927 75 65
Source: WVE Baseline Survey (2001) and Own Sampling

36
4.3 Methods of Data Analysis

Both descriptive and econometric analyses were employed to meet the specific objectives of

the study. By applying descriptive statistics, one may compare and contrast different

categories of sample units with respect to the desired characters so as to draw some important

conclusions. In this study, descriptive statistics such as mean, frequency, percentage, t-test,

chi-square were used to analyse the collected data.

4.3.1 Cobb-Douglas Production Function

Moreover, CD production function was used to examine the agricultural productivity

differences between the female and male headed households. According to Gujarati (1995),

the generalized form of the CD production function can be specified as:

Y = AX 1B1 X 2B 2 X 3B3 .......... ......... X nBn eUi --------------------------------------------------- (2)

Where, Y is gross value of farm outputs in Birr per ha, Xi’s are explanatory variables such as

land size, livestock holding, education level, fertilizer use, herbicides use, male and female

labour; Bi ’s are coefficients or elasticities of output and indicate how strongly each input

affects output; A is efficiency parameter and represents the level/state of technology and U i

is disturbance term.

Since the CD production function is a power function, it is impossible to directly use the

Ordinary Least Square (OLS) Method. Therefore, logarithmic transformation was made to

obtain its linear form and to estimate the parameters. In this study, the natural logarithm was

employed.

37
To examine whether the production functions of male headed (MHH) and female headed

households (FHH) are different from each other, Equation (2) was estimated separately for

MHH and FHH. Moreover, pooled data set without and with a dummy gender variable was

estimated. The respective transformed models are shown as follows:

Production function for MHH:

ln Y = ln A + B ln X +B ln X +B ln X + ....... + B ln X + U m - (3)
m m 1m 1m 2m 2m 3m 3m 7m 7m

Production function for FHH:

ln Y = ln A + B ln X + B ln X + B ln X + ........ + B ln X + U f --- (4)

f f 1f 1f 2f 2f 3f 3f 7f 7f

Production function using pooled data:

ln Yp = ln A + B ln X + B ln X + B ln X + .... + B ln X + U p --------(5)
p 1p 1p 2p 2p 3p 3p 7p 7p

Production function using pooled data with dummy gender variable:

ln Yp = ln A + B ln X + B ln X + B ln X + .... + B ln X + DG + U p ---(6)
p 1p 1p 2p 2p 3p 3p 7p 7p

Where, m =MHH, f =FHH, p =Pooled data set, G =Gender dummy variable (G=1 for

MHH; G=0 otherwise) and D is the regression coefficient for the dummy variable and it

indicates gender differences in technical efficiency. Bim , Bif and B ip (i=1, 2, ---,7) are output

elasticities of ith input under MHH, FHH and pooled data sets, respectively.

Moreover, MVP of inputs were computed from the coefficients of the regression. MVP of a

factor is the additional return from adding one more unit of that factor, holding all other

inputs constant. Comparing the MVP of a factor with the prevailing factor cost (opportunity

cost) shades some light on the efficiency of resources use. MVP computed at the mean value

of inputs may be used to indicate whether disequilibrium in resource use is big or small

38
(Ellis, 1988).

The MVP of the factor can be computed as follows (Ellis, 1988):

Y
MVP = bi * ----------------------------------------------------------------------- (7)
Xi

Where bi is the regression coefficient (output elasticity), Y is the gross value of farm output

(geometric mean) and X i is the geometric mean value for factor i.

Finally, Oaxaca decomposition model of the productivity differential between male and

female farmers was used to decompose the productivity difference (Oaxaca, 1973). Although

this approach was used to decompose the wage gap, it can also be applied to decompose

productivity difference between, say, men and women farmers (Quisumbing, 1995). The

decomposition model adopted was presented as follows:

Y   X 
Ln  m  = [( Bim − Bif ) LnX if ] + BimLn ( im ) ------------------------------------------------- (8)
 Yf   X if 

Where Ym and Y f represent mean output (Geometric mean) of males and females

respectively, X im and X if are geometric mean levels of inputs of male and female, Bim and

Bif are estimated output elasticities of male and female headed households as defined earlier.

The model decomposes the overall average male-female output gap into the portion due to

differences in the technical efficiency and the portion attributable to differences in input

endowments. In other words, the first bracketed expression on the right hand side is a

measure of change in output due to shift in output elasticities of the production functions.

39
The second bracketed term is a measure of difference in output due to difference in volume

of input use per hectare.

4.3.2 Estimation Technique and Testing Procedures

The estimation technique employed in this study was OLS. The OLS has very attractive

statistical properties that have made it one of the most powerful and popular methods of

regression analysis including linearity, unbiasedness and minimum variance (Koutsoyiannis,

1997). All the production functions were estimated separately using OLS techniques with the

help of Statistical Package for Social Science (SPSS) computer software.

Before estimation of the models, the severity of multicollinearity (the linear relationship)

among continuous explanatory variables was checked by computing the Variance Inflation

Factor (VIF). VIF shows how the variance of an estimator is inflated by the presence of

multicollinearity. Following Gujarati (1995), the VIFi is given as:

 1 
VIFi =   ----------------------------------------------------------------------- (9)
 1 − Ri
2


Where R2 i is the coefficient of determination that is obtained when the continuous

explanatory variable is regressed against all the other explanatory variables. As R2 i

approaches 1, the VIF approaches infinity. That is, as the existence of collinearity increases,

the variance of the estimator increases and in the limit it can be infinity. If there is no

collinearity between regressors, the value of VIF will be 1. As rule of thumb, if VIF of a

variable exceeds 10, that variable is said to be highly collinear (Gujarati, 1995).

40
In order to test the homogeneity between the parameters of the production functions, the

Chow’s test was performed. This examines whether the male and female headed household

functions differed significantly due to shift in the intercept or/and due to change in the slope

of the functions.

The major steps adopted in computing Chow’s F-test was as follows: Firstly, estimation of

production functions for male and female headed households was made to obtain their

residual sum of squares, say, Σe21 and Σe22 with n1 -k and n2 -k degrees of freedom,

respectively. Where, n1 and n2 are the number of observations in male and female headed

households and k is the number of parameters including the constant.

Secondly, the pooled function (Equation 5) was run to obtain residual sum of square, Σe2p ,

with n1+ n2 -k degree of freedom.

Thirdly, compute F as

(∑ e − (∑ e + ∑ e ) )/ K
2 2 2

F=
p 1 2

(∑ e + ∑ e )/( n + n − 2K )
2 2
---------------------------------------------------------(10)
1 2 1 2

In the last step, the computed ‘F’ ratio was compared with the theoretical value of ‘F’ with k

and (n1+ n2 -2k) degrees of freedom and reject the hypothesis that the separate regression are

the same if the calculated F value is greater than the tabulated one (Koutsoyiannis, 1997).

The significance of the calculated Chow’s F-statistic implies that there is a shift in the

production function between male and female headed. If there is a change in the parameters

of the two functions, it can be said that the function has undergone a structural change

41
(Koutsoyiannis, 1997). Structural change may mean that the two intercepts are different or

the two slopes are different, or both the intercept and the slopes are different. However, the

above test does not tell whether the shift is in scale parameters (intercept term) or slope

parameter (elasticity coefficient).

The significance of the coefficient for the gender dummy (D) in Equation (6) implies that

there is a shift in the intercept term. If, on the other hand, the coefficient is insignifica nt, then

both production function have the same intercept term. Again, in order to check whether

there is a shift in the slope parameters, the Chow’s F-value was computed using residual sum

of squares for the pooled production function with a dummy gender (Equation 6) instead of

that of Equation (5).

The Chow test does not explicitly show which slope coefficient or intercept term is different

from each other in the two production functions, i.e. MHH and FHH. Therefore, in order to

check the equality between the coefficients of both the production functions, the log- linear

transform of Equation (2) was estimated with both intercept and slope dummies as follows:

lnYp =lnC+C1lnX1p +C2 lnX1p +......+C7 lnX7p +C8Gi +C9(Di lnX1p) +...+C15(Di lnX7))+Up- (11)

Where G i is a dummy with a value of one for MHH and zero otherwise ; other variables being

as defined earlier. If all or some of the coefficients of the slope dummies (C9 to C15 ) are

positive and significant, then the output elasticities of MHH with respect to that particular

input is greater than that of FHH, otherwise they are the same.

42
CHAPTER FIVE: RESULTS AND DISCUSSIONS

5.1 Descriptive Analysis

This section discusses the nature of the socio-economic and demographic characteristics of

the households that affect men’s and women’s productivity in agriculture using descriptive

statistics. As discussed earlier in the methodology part, the study was based on cross

sectional data obtained from 140 households of which 75 were male headed and 65 were

female headed (dejure). The discussions in this section mainly compare the two household

groups.

5.1.1 Household Characteristics

About 98% of the sample households are Oromo while the rest i.e. 2% of the households

accounted for Guraghe and Amhara. In terms of religion, the result of this survey shows that,

about 89.3%, 9.3% and 1.4% of MHH were Orthodox, Protestant and Muslims, respectively.

The corresponding figures for FHH were 89.2%, 9.2% and 1.6%, respectively.

Total number of family members of the sample households was about 692 out of which

49.7% were male and 50.3% were female. This figure was consistent with secondary data

obtained from CSA (2002) which indicated that male constitute about 49.2% of the total

population of the district.

The FHH has about 39.4% of the total family members in the sample. Out of these, male

members accounted for 43.2%. And when we compare the number of male in female headed

households with the total, the share was only 34% while the remaining 66% belongs to the

MHH. The average household size was 5.6 and 4.2 persons with a range of 2 to 10 and 1 to 9

for male and female headed households, respectively. As the t-test shows the mean difference

43
was statistically significant (t=4.7, p=0.01). This implies that male headed households have

larger family size than female headed households.

In addition to the unscaled family size, adult equivalent (AE) that takes care of age and sex of

individuals was also computed (Table 3). To compute household size in AE, standard

conversion factors given in Appendix 3 was used. Accordingly, the mean family size was 4.4

and 3.2 in MHH and FHH, respectively, which is statistically different (t=4.62). Saito et al.

(1994) also indicated that FHH had relatively lower family size as compared to MHH in their

study conducted in Kenya and Nigeria. The result is also consistent with that of Dejene

(1994) and Addis et al. (2000).

Table 3. Average Family Size of the Households in AE

Sex MHH (N=75) FHH (N=65) Total (N=140) t-value

Male family member 2.6 1.5 2.1 5.2 *

Female family member 1.8 1.7 1.8 0.7

Total 4.4 3.2 3.8 4.62*

Source: Own survey (2003) NB: *=Significant at 1% probability level

The age structure of sample households shows that the average age of male household heads

was 46.6 years compared to 47.8 years for female heads with the minimum and the

maximum age of 17 and 80 years for MHH and 21 and 75 years for FHH, respectively. This

difference was not statistically significant (t=-0.483, p=0.63). Regarding economically active

family members (15 to 65 years), the MHH had larger economically active members (2.4

AE) than FHH (1.7 AE), significantly different at 1% probability level (t=4.02). The age

distribution in both groups shows that the economically active age groups constitute the

44
largest share of the family members (Table 4).

Table 4. Average Family Size by Age Group and Sex of Household Head (AE)
Age group MHH (N=75) FHH (N=65) All cases (N=140) t-value

<5 year 0.5 0.3 0.4 2.53**

5-14 year 1.3 1.2 1.2 0.67

15-65year 2.4 1.7 2.1 4.02*

>65year 0.1 0.1 0.1 1.19

Total 4.4 3.2 3.8 4.62*

Source: Own survey (2003) NB: *, **=Significant at 1% and 5% probability level respectively

The educational status indicates that about 66% of the FHH were illiterate; about 13.8%

attended literacy classes while around 20.2% had primary education. In contrast, about

45.3% of MHH were illiterate, 12% attended literacy classes while 22.6% and 20.1% had

primary education and secondary education, respectively. On the average, head of male

house holds attended 2.73 years of schooling while that of female heads attended 0.75 years.

This shows that there is a significant difference (t=4.56, p=0.01) in terms of access to formal

education between MHH and FHH (Table 5).

Furthermore, the average number of years of schooling for MHH family members was about

4 years while it was 2.8 years for FHH, which is significant at 5% probability level (t=2.25).

On the other hand, out of the 146 school age children in MHH about 71% attended school

while out of 134 school children about 52% attended school in FHH. This also indicates that

children in FHH had less access to education as compared to that of children in MHH. In

general, FHH had less access to formal education as compared to MHH in the study area.

45
The literacy rate in the district, according to data from the education office of the district, is

about 30%.

Table 5. Literacy Status of Household Head (%)

Literacy MHH (N=75) FHH (N=65) All cases (N=140)
Illiterate 45.3 66 55
Read and write 12 13.8 12.9
Grade 1-6 22.6 20.2 21.4
Grade 7-8 10.6 - 5.7
Above Grade 8 9.5 - 5
χ2 =15.952 p=0.003
Mean years of schooling 2.73 0.75 1.73
Source: Own survey (2003)

In most sub-Saharan Africa, the adult literacy rate of men is almost twice that of women and

the enrolment of boys is almost twice as that of girls in secondary school. Gender based

educational discrepancies tend to be greater in countries where incomes are lower (Saito et

al., 1994; Saito and Surpling, 1992). Studies conducted by Addis et al. (2000), Tiruwork

(1998) and Dejene (1994) in Ethiopia also show that FHH have less access to formal

education.

Regarding marital status of the sample respondents, about 75.4% of FHH were widow and

97.3% of MHH were married (Table 6). The difference is quite significant as shown by the

chi-square value of 133.97.

46
Table 6. Marital Status of the Sample Households (%)
Marital status MHH (N=75) FHH (N=65) All cases N=140)
Married 97.3 - 52.2
Widow - 75.4 35
Unmarried 2.7 9.2 5.7
Separated/Divorced - 15.4 7.1
Total 100 100 100

chi-square=133.97 P=0.00
Source: Own survey (2003)

As shown in Table 7, the major reasons behind the cause for females to be head of the

household were death of husbands (75.4%), divorce (15.4%) and death of their father and

mother (9.2%). Of the respondents (10 FHH) who reported divorce as causing female

headship, 8 of them attributed divorce to economic problems and personal conflict while the

other 2 attributed to polygamy. The economic problem implies that in most cases when a

husband lacks resources to sustain a household, the husband or the wife opts for separation.

Table 7. Reasons for Being Female Headed (dejure)

Reasons Number of female headed %
Husband died (Widow) 49 75.4
Divorce 10 15.4
Death of father and mother 6 9.2
Total 65 100
Source: Own survey (2003)

Baseline survey conducted by World Vision International Ethiopia (WVE) in 2001 shows

that, on average, 27% of the households were found to be female headed including defacto

while the rest 73% were male headed. As to formation of defacto headed households,

migration of male to other areas in search of job opportunities is the major cause for the

47
women to lead the household. This is usually for certain seasons but sometimes for longer

period leaving women to subsist themselves and their children. According to the information

from the district council, some of them migrate to other places like Jima (coffee picking),

Dire Dawa (trade), Adola (Gold extraction), etc. The number of migrants from the district is

relatively higher in towns like Dire Dawa where they even have formed an association called

‘Chebo’ to promote mutual help.

5.1.2 Farm Characteristics

5.1.2.1 Land Use Pattern

The total area of land owned by the sample farmers was about 127 ha with the average of

0.91ha per household. Compared to the regional average, which was 1.13 ha, landholding

under different uses was small in the study area. As indicated in Table 8, the average land

owned by MHH and FHH was 1.06 and 0.72 ha respectively, which is statistically different

at 1% probability level (t= 3.35). This indicates that landholdings of the households headed

by women are smaller by about 47% than that of the male headed. According to the

information obtained from the district council, scarcity of land is the major cause for

migration of males to other areas.

Some studies conducted in Africa have showed similar results. For example, in Nigeria the

FHH surveyed had only one third of the land holding of MHH and in Kenya the holding of

FHH were 65% of MHH (Saito et al., 1994). Other studies also indicate that, in general, FHH

in sub-Sahara Africa have tended to cultivate smaller plots of land than MHH (Ndiyo and

Urassa, 2002).

As shown in Table 8, about 72% and 73% of the total land owned was put under cultivation

48
of different crops by male and female headed households, respectively, while the rest was

allocated for grazing and trees/woods and some are waste land.

Table 8. Land Use Pattern of the Households (Mean in ha)

Activities MHH (N=75) FHH (N=65) All cases (N=140) t-value
Own cultivated land⊗ 0.77 0.52 0.65 3.36 *
Grazing land 0.20 0.11 0.16 3.16 *
Forest/wood land 0.06 0.05 0.06 0.88
Waste land 0.03 0.04 0.04 -0.28
Total 1.06 0.72 0.91 3.35 *
Source: Own survey (2003)
NB: *=Significant at 1% probability level ⊗ excluding rented, borrowed and shared land

As it can be seen in the Figure 4, larger proportion of the households (27%) have land

holding between 0.76 and 1 ha. About 62.6% of MHH and 72.4% of FHH have landholding

of less than one hectare.

Figure 4. Distribution of Land Size (ha)

35

30

25
Percentage

20 MHH
FHH
15 All cases

10

0
<0.25 0.26-0.5 0.51-0.75 0.76-1.00 1.01-1.50 1.51-2.00 >2.01

Land size(ha)

Source: Own survey (2003)

49
Although there were other crops produced in the study area such as field pea, chickpea, noug,

lentile and linseed, only the major crops are discussed here. The major crops cultivated in the

area are teff, barely, wheat, sorghum, bean, maize and enset. The average total land areas

under these crops were about 1.01 and 0.66 ha for MHH and FHH, respectively. In the total

cropped area, teff is the major crop for both MHH and FHH followed by wheat. The size of

land allocated to the major crops of the area is presented in Table 9.

Table 9. Cropping Pattern of Sample Farmers

Crops MHH (N=75) FHH (N=65)
%∂ Area (ha) % Area t-value (Area) χ2
Teff 64 0.31 58 0.22 2.02** 0.7
**
Barely 33 0.11 22 0.05 2.35 2.4
*
Wheat 71 0.26 49 0.13 3.51 5.8*
Sorghum 15 0.03 12 0.03 0.268 1.6
Bean 11 0.03 26 0.05 -1.41 2.6
Maize 47 0.12 42 0.10 0.91 0.1
Enset 95 0.15 89 0.08 3.73* 0.7
Average 1.01 0.66 4.73*
Source: Own survey (2003) NB: *,**=Significant at 1 and 5% probability level respectively
∂=indicates the percentage of farmers cultivating the major crops listed in each group of the sample.

About 97% of MHH and FHH had farming as their primary occupation in the study area with

the same number of farming experience i.e. 30 years each. People in the area rely on rainfall

for their crop production. As stated in Table 9, the chi-square test shows that there is no

difference between the two groups of household regarding cultivation of crops. This may

show that there is no gender specialization with these crops in the study area. In other words,

there is no gender assignment of crops among this community. Enset is the only perennial

crop in the area and produced both for cash generating and consumption purposes. About 95

50
% of MHH and 89% of FHH are growing enset (χ 2 =31.28, p=0.31). On average, the MHH

have about 154 seedlings of enset on their farm while FHH have about 76-enset plant.

Moreover, it was found that the fertility of land in Wenchi varies from productive plots to

severely eroded plots. Only 43% of MHH and 20% of FHH indicated that their land is fertile

while the rest are medium and/or eroded, according to farmers’ judgement. It was assumed

that the distribution of land is homogeneous among farmers but eac h farmer has got each

type of land. To avoid soil fertility losses, the farmers practice various soil and water

conservation techniques like terracing, cut off drain and tree planting. The correlation

coefficient between yield of crops and area under soil and water conservation activities was

about 0.2, which is statistically significant at 1% probability level.

5.1.2.2 Livestock Holding

Livestock rearing is another economic activity in which the majority of MHH (93%) and

FHH (74%) participate along with crop production. The remaining 7% of MHH and 26% of

FHH reported to have no livestock. The major livestock reared by the respondents are cattle,

sheep, goat, donkey, horse and chicken. In order to convert the number of different livestock

the farmers own into standard unit, each animal was converted into a Tropical Livestock Unit

(TLU), which is measured as equivalent of a live weight of 250 kg. Livestock population

number is converted into TLU using conversion factors given in Appendix 2.

The mean livestock size owned by the sample farmers was 2.19 TLU. Comparison of the

livestock ownership between MHH and FHH shows that MHH on average own 2.69 TLU

while FHH own 1.61 TLU, which is significantly different at 1% probability level (t=3.38).

This implies tha t MHH were relatively wealthier than FHH, since livestock is considered as

one of the indicators of wealth status in the study area.

51
Table 10. Livestock Ownership by Gender (TLU)
Type MHH (N=75) FHH (N=65) All cases (N=140) t-value
Oxen 0.85 0.50 0.69 2.59**

Cattle ω 1.51 0.99 1.27 2.58**

Sheep/Goat 0.08 0.05 0.07 1.30
Chicken 0.02 0.01 0.02 1.99***
Pack animal (horse 0.23 0.06 0.14 2.74**
and donkey)
Total 2.69 1.61 2.19 3.38*
Source: Own survey (2003) NB: *,**,***=Significant at 1% ,5% and 10% probability level
respectively ω=Includes cow, calf, heifers and bulls
5.1.2.3 Labour Utilization

Agricultural labour among the community can be divided into three: first, family labour,

which consists of the labour force of all able persons residing within the household. The

second is hired labour, which simply engages in the agricultural fieldwork of the household

and paid at the end of the working day. The third form of agricultural labour that a household

organizes is debo3 . A household can organize work party such as debo when it finds that

particular agricultural task cannot be managed by the household labour alone.

According to the result of the survey, FHH has smaller family size and have fewer number of

adults for farming than MHH, which is statistically different at 1% probability level (t=4.84).

MHH have on the average about 1.54 men adult while FHH have 0.78 men. The average

number of working days in a month, though it differs from one month to another and from

one area to another, was about 17 days for MHH and 14 days for FHH. People in the area

highly celebrate holidays during the months. They do not participate in any activity

especially on the farm during holidays. This may be one of the reasons for the labour

shortage facing the farmers in the area. Respondents were also asked to quantify the amount

of labour they put on major activities of crops production. Accordingly, the average man-

52
days of family labour used to produce crops on a hectare of land was reported to be 147 and

88 for MHH and FHH, respectively (t=3.4, p=0.001). On the other hand, the average man-

days of hired labour for MHH and FHH was 10 and 12, respectively, which is not

significantly different (Table 11).

Table 11. Sources of Labour Use in Man-days per ha

Source MHH (N=75) FHH (N=65) All cases (N=140) t-value
Male family 107 42 77 4.43 *
Female family 40 46 43 -0.46
Hired labour 10 12 11 -0.91
Total 157 100 131 3.04 *
Source: Own Survey (2003) NB: *,**=Significant at 1% probability level

About 41% of MHH and 68% of FHH reported that labour shortage constrains effective

undertaking of their agricultural activities. The chi-square test shows the existence of

significant difference in labour shortage for different activities between MHH and FHH

(χ2 =9.73, p=0.002). This means that FHH have more constrained by labour as compared to

the MHH. The average wage rate and working hours for all activities were 6 Birr and 8 hours

per day, respectively.

The major strategies used by MHH towards overcoming labour shortage were hiring labour

(32%), assistance from relatives (29%) and through social support such as debo and

dado4 (39%). The corresponding figures for FHH were 34, 34 and 14%, respectively and

about 18% of the FHH were not able to overcome the problem at all (Table 12).

3
Debo is a labour pooling system where a farmer requests his neighbours or relatives for help on a particular
day without an agreement to perform their activity in return.
53 The farmer provides the participants with food and
drinks on the farm and at home.
4
Dado is a system of labour sharing where farmers make a mutual agreement to perform their activity (planting, weeding,
harvesting, transporting activities) turn by turn in a rotation basis. The number ranges between six to nine in most cases.
Table 12. Strategies Used to Overcome Labour Shortage (%)
Strategies MHH (N=75) FHH (N=65)
Hiring labour 32 34
Assistance from relatives 29 34
Social support 39 14
Not able to overcome - 18
Total 100 100
chi-square=12.59 p=0.027
Source: Own survey (2003)

5.1.2.4 Use of Fertilizer, Herbicides and Improved Seed

Commercial inputs are widely used by farmers in the study area. About 88% of MHH and

54% of FHH applied commercial fertilizer while 15% of MHH and only 5% of FHH used

improved seeds of maize, wheat and teff. The average amount of fertilizer used by MHH and

FHH was 120 and 51 kg/ha, respectively (Table 13). These figures show that there were

significant differences in the use of fertilizer at 1% probability level (t=5.14) between the two

groups. Fertilizer application is generally far below the widely recommended rate of 100 kg

of DAP and 50 kg of Urea per hectare.

With regard to herbicides, about 48% of MHH and 26% of FHH used herbicides to control

weeds. The mean amount used by MHH and FHH was about 0.26 and 0.19 litre respectively.

This amount is below the recommended rate of one litre per hectare, according to the

information from the agriculture office of the district. The respondents pointed out that Mujja

(Snowdenia ploystachya), Sinnara (Avena species) and Tuffo are the major weed species in

the area

54
Table 13. Mean of Quantity Applied and Cost of Agricultural Inputs Applied

MHH (N=75) FHH (N=65)

Type Quantity Cost Quantity Cost t-value
(Kg/ha) (Birr/ha) (Kg/ha) (Birr/ha)
DAP 61 152 25 63 5.19*
Urea 59 118 26 53 4.63*
Improved seed 15 68 5 23 1.8 ***
Herbicides 0.26 12 0.19 9 1.14
Source: Own survey (2003) NB: *,***=Significant at 1% and 10% probability level.

5.1.3 Crop Yield

The average yield of teff for MHH and FHH was about 5.1 and 4.3 quintals 5 per ha,

respectively, which was statistically significant at 10% (t=1.75). The average yields of

barely, wheat and maize were about 9.1, 7.9, 5.1 Qt per ha for MHH, respectively and for

FHH the respective yield of these crops was 7.3, 6.4 and 4.3 Qt per ha. This indicates that

MHH had higher yield per ha of land compared to FHH (Table 14).

Table 14.Yield of Major Crops by Gender (Qt/ha)

Crops MHH (N=75) FHH (N=65) t-value
Teff 5.1 4.25 1.75 ***
Barely 9.08 7.3 1.19
Wheat 7.85 6.4 1.98 ***
Sorghum 9.2 4.1 7.0*
Bean 6.56 4.94 1.31
Maize 5.1 4.25 1.75 ***
Source: Own survey (2003) NB: *, ***=Significant at 1 and 10% probability level, respectively

55

5
One quintal is equal to 100 kg
5.1.4 Gender Based Division of Labour

Generally speaking, women and men in Wenchi have clearly separate labour roles to play.

The main criteria for the division of labour in the area are age and sex. Women are entirely

responsible for reproductive activities in and around the household while men do most of the

work on farm or work for wages (Dejene, 1994; Tiruwork, 1998).

Heads of the households were asked about the major activities performed by all the family

members to see whether the agricultural production system is gender oriented or not. The

major activities (productive, reproductive and community works) undertaken by the

respondents were ranked according to the number of men and women family members

participated in the activities.

Both men and women take active part in production of all crops produced in the area. In the

production of these crops, women were engaged in all agricultural work except ploughing

using oxen, sowing and enset planting. On the other hand, men participated in all of the

activities starting from land preparation to marketing. The activity profile of sampled

households is summarized in Table15.

56
Table 15. Proportion of Male and Female Family Member Participated (activity profile) 
Activities MHH (N=75) FHH (N=65)

Male Female t-value Male Female t-value

Land clear ing 75 31.1 6.79* 64.6 47.4 -0.23

Land ploughing 73.6 0.0 - 46.3 0.0 -

Sowing 55.4 0.0 - 40.2 0.0 -

Weeding 69.6 56.1 1.51 43.9 60.3 -4.28*

Harvesting 76.4 18.2 8.96* 54.9 49.1 -1.26

Threshing 81.8 8.3 11.11 * 53.7 25.9 1.33

Transporting 79.7 45.5 4.59* 37.8 64.7 -4.41*

Marketing 41.9 58.3 -1.59 24.4 70.7 -8.63*

Enset processing 0.0 77.3 - 0.0 76.7 -

Enset planting 75.7 0.0 - 79.3 0.0 -

Vegetable gardening 21.6 38.6 -3.14* 11.0 19.0 -2.08 **

Livestock herding 60.8 18.2 8.77* 79.3 31.9 3.28*

Fetching water 1.4 84.8 -21.76 * 9.8 90.5 -13.86 *

Fire wood collection 6.8 78.8 -15.11 * 11.0 84.5 -11.89 *

Cleaning house 0.0 84.8 - 2.4 88.8 -16.25 *

Food preparation 0.0 79.5 - 0.0 84.5 -

Milking cow 0.0 18.9 - 0.0 6.9 -

Child care 10.1 40.9 -5.79* 1.2 13.8 -3.21*

Grain mill 0.0 80.3 - 2.4 75.9 -12.29 *

Source: Own survey (2003). NB: *, **=Significant at 1 and 5% probability level, respectively
 it is calculated by dividing the number of household members participated in each activity with the total
household members above age of 10 years.

57
Land preparation is carried out using oxen as draught animal and some times they use

diqdiqa6 , especially those in dega agro-climatic zone. Ploughing is entire ly men’s activity.

However, women may help in clearing the land and in softening of the soil. Women would

never try ploughing using oxen. There is a belief that “yoo dubartiin qottee, roobni hin

roobuu” means if women cultivate, there will be no rainfall. Sowing and enset planting are

also activities carried out only by men. These activities are culturally banned and strictly

cotempted for signal of a disaster to the community if women do them.

Other agricultural activities are shared with women. Digging and weeding is a responsibility

of both men and women and is done either alone or together. But not all crops need both

digging and weeding. Digging is mainly for maize followed by weeding. In most cases

weeding is one time activity. Weeding is unavoidable for maize and is a usual task for teff.

According to the result of this study, about 70% of men and 56% of women family members

in MHH and about 44% of men and 60% of women family members in FHH participated in

weeding (Table 15).

Next to weeding and digging, harvesting, collecting and threshing in order are relatively

common tasks of women in the area. Harvesting involves different tasks for different crops:

reaping for teff, wheat and barely; uprooting of haricot bean; and stripping off the cob for

maize. Preparing the threshing ground is a woman’s task while threshing is done by men

(majority) and women (few especially in FHH where there is no male). Collecting and

transporting is frequently done by women but using pack animals men do transport grain

from field to home.

58

6
Farm implement used for digging of soil
With regard to livestock production, men are responsible for herding, usually assisted by

boys. They also feed cattle, track animals for drinking water sources and vaccination centres.

It is only in the absence of male that women take the responsibility of looking after the

livestock. Women do milking cows, clean cattle barns, buying salts for cattle and taking care

of sick animals. In general, as women usually look after milking cows, men give care to

oxen.

Food preparation including grinding of grains, preparing coffee, etc are mainly done by

women. They are also responsible for cleaning the house, fuel collection (either cow dung or

fire wood) and water fetching. Tending of children is mainly the duty of women, too.

According to the respondents, enset processing consumes most of the women’s time and is

the hardest work in the area for women. The routine tasks performed in extracting an edible

product from enset are pit digging, cutting and uprooting, decorticating (separating the edible

part of pseudo stem from the fibre), pulverizing (chopping the corm into pieces), burying and

fermenting, squeezing the fermented mixture, chopping to minimize the fibrous part of the

plant and sifting the mixed enset product using home made sieve. All these activities are only

done by women.

Husband and wife go to the market alone or together but in most cases they go together with

people from their neighbourhood. The average distance they travel from the nearest market is

about 2 km whereas the furthest market is about 9.2 km. Men market the high value assets

such as livestock and larger quantities of grain while women sell or buy smaller quantities.

59
In general, the result of this study shows that women work on the average for 13 hours a day

and the working hours increase during the farming season. On the other hand, men on

average work for 9 hours. Paired samples t- test indicates that there was a significant

difference between working hours of the two groups at 1% probability level (t=-16.9).

From the above discussions, it can be concluded that some of the agricultural activities

undertaken in the study area are gender oriented; especially enset planting, land ploughing

using oxen and sowing are only undertaken by men while enset processing, food preparation

and other domestic works are done by women.

5.1.5 Sources of Farm Income

Sales of crops, livestock and off-farm activities are the major cash income sources for the

households in the study area. About 76% of MHH and 52% of FHH reported that they earned

cash income from sales of crops whereas about 43% of MHH and 49% of FHH earned cash

income from sales of livestock and /or livestock by-products. The average cash income from

different crops was about Birr 345 for MHH and Birr 150 for FHH, the difference is

significant at 1% probability level. Income from crop sale constitutes the highest proportion

for MHH while it is income from livestock sales for FHH as indicated in Table 16.

Table 16. Source of Cash Income by Gender

Income sources MHH (N=75) FHH (N=65) t-value
% Mean (Birr) % Mean (Birr)
(Birr)
Sale of crops 76 345 52 150 5.38 *
Sale of livestock and/or their products 43 151 49 169 -0.43
Off-farm income 4 35 23 82 -1.39
Total 531 401 2.03 **
Source: Own survey (2003) NB: *, **=Significant at 1 and 5% probability level, respectively

60
As pointed out earlier, agriculture in Wenchi does not satisfy the basic needs of the people.

Therefore, people have to acquire alternative sources of income. About 4% of MHH and 23%

of FHH were engaged in different off-farm activities such as petty trade, working as daily

labourer, spinning and weaving, which is statistically different between the two groups

(χ2 =11.31, p=0.001). This indicates that FHH were highly involved in off- farm activities as

compared to MHH. Some of the daily labourers (mainly men) go to urban centres (Welliso)

to take up all sorts of daily works and some (mainly women) work on the farms of richer

farmers.

5.1.6 Access to and Control Over Productive Resources

The distinction between access and control is based on the conceptual separation of the use

of resources and services and the benefits derived from their use. According to CISP (1997),

access can be defined as “the opportunity to make use of something” and control as “the

ability to decide the use of something”. The distinction is important because access does not

necessarily imply that an individual controls a particular resource or the benefits from its

utilization. For example, a woman may earn an income from productive activities but have

no control over how it is spent. Generally, access is a precondition for control.

Below are brief discussions on access and control over major productive resources, both

physical and institutional, by male and female ho useholds.

5.1.6.1 Land

Obviously, land in rural areas is a very important means of production. It does not only play

a central role in producing crops and livestock but also it is a security for getting credit

services. Moreover, access to land offers a privilege to get access to agricultural extension

61
services and new agricultural technologies. But due to customary laws, culture and tradition,

access to land is closely related to gender. Men are registered as the owner of the household

land. They pay land use taxes in the name of male. Men household heads are privileged to

have all rights and duties concerning the household land. Wives have no right to claim the

household land unless their husbands die. Even after husband’s death, his name remains on

the PA land use tax book. Widows rather take-over the right and duties of acting on behalf of

their husbands. They continue paying land use taxes in husbands’ name. If any controversy

over the land occurs, widows go to court in the name of the deceased husba nds. In fact,

following the death of her husband a woman soon becomes the household head and assumes

the responsibility of administering the household and all properties including land.

According to information obtained from administration council of the district, since 1975

land distribution had not been taken place in the district. As a result, land is obtained through

different means such as inheritance upon death of one’s father or husbands and through

marriage gift. In principle, women in Ethiopia are entitled to land. The land reform

legislation has provided men and women equal rights of owning land. According to

proclamation No 31, 1975 (Article 4), “without differentiation of the sexes, any person who

is willing to personally cultivate land shall be allotted rural land sufficient for his/her

maintenance and that of his/her family”. This provision clearly shows that both women and

men have equal rights to own land.

However, despite the provision of the law, women in the study area have never gained rights

to own land because of socio cultural systems. Therefore, land has been and is still

considered to be the sole property of male members and is transferred from generation to

generation on the basis of rule of decent through male. This is the accepted rule both by men

62
and women members of the society as control of men over land is along established tradition

to which the society at large still adheres. This emanates from the belief that after marriage a

daughter would start a new life with her husband and hence potentially belongs to others,

while the son follows the footstep of his father and is believed to properly maintain and

manage the whole family even after the father passes away.

Upon marriage the husband and wife pool their plots and cultivate as a single unit. However,

according to the respondents, upon divorce women may find it difficult to get back the land

they owned before marriage. In most cases, when this condition happens men desire to retain

control of it. Moreover, women’s lack of legal capacity prevents women from acting

independently in matters of property acquisition or transfer and in all other legal transactions

where customary or legally responsible male’s approval is needed. Thus, women in the study

area are dependent upon their fathers, husbands or brothers.

As it can be seen from Table 17, the land owned by respondents was acquired through

inheritance, provided by family and government. About 16.7% of men and only 3.6% of

women have got land from their families. A higher percentage of FHH (85.7%) obtained

their land through inheritance while about 48.6% of MHH from government during land

distribution made in 1975.

Table 17. Proportion of Households Acquiring Land From Different Sources (%)
Sources MHH (N=75) FHH (N=65) All cases (N=140)
Given by PA 48.6 10.7 32
Inheritance 34.7 85.7 57
Given by family 16.7 3.6 11
chi-square=33.42 p=0.01
Source: Own survey (2003)

63
The land used for cultivation can also be obtained through land transaction. Accordingly,

about 96% of MHH and 86% of FHH have own land. That means about 14% of FHH and 4%

of MHH have no owned land. The others access land through renting and borrowing.

Moreover, about 3% of FHH have no access to land at all while all of the sample MHH have

access to land either from rented, borrowed or shared (Table 18).

Table 18. Type of Land Access by Gender

Access MHH (N=75) FHH (N=65) t-value
% of HH Mean (ha) % of HH Mean (ha) (Mean)

Owned 96 1.06 86 0.72 3.35 *

Rented 29 0.17 19 0.10 1.53
Borrowed 1 0.002 5 0.009 -1.42
Shared 8 0.03 20 0.13 -2.27**
No access - - 3.1 - -
Source: Own survey (2003) NB:*,**=Significant at 1 and 5% probability level, respectively

5.1.6.2 Oxen

Like in most parts of Ethiopia, in the study area, draught animals are used as a source of

power for farming. On the average, MHH had about 0.85 oxen while FHH had 0.5 oxen,

which was significant at 5% probability level (t=2.59). The survey indicates that about 45%

of MHH and 55% of FHH did not possess any oxen, while those owning only one ox

constitute 27% of MHH and 39% of FHH. This shows that FHH has less access to draught

oxen as compared to FHH in the area.

The most widely used method of overcoming shortage of oxen was exchange of labour for

oxen, pairing oxen with others, borrowing oxen from relatives and hiring oxen. In some cases

women who have no oxen have their land ploughed by giving services such as weeding,

64
clearing the land for a week for the owner of the oxen. Similarly, if FHH has a male labourer,

he can work for those who own oxen in exchange for the use of the oxen to cultivate the

land. An individual should plough two days for the oxen owner in order to use a pair of oxen

on his land for one day.

On the other hand, those who have no adult male labour are forced to give out land to share

croppers. Share cropping (Hirta) is practiced if she/he has land and not able to cultivate

because of shortage of labour or other inputs, he/she provides the land to somebody and

shares the output equally. From the result of this survey, it can be seen that about 20% of

FHH and 8% of MHH gave out their land for sharecropping (Table 19). Sharecropping out is

a process whereby the land owner and the partner make mutual agreement to give a certain

porion and/or whole of one’s farm plot for a specified period of time by specifying the input

contribution and deciding on the share of the output.

As regarding to gender based ownership of oxen, due to cultural reasons in all the study area,

draught animals are considered as the property of men. Even in FHH, oxen are considered as

the property of the male members.

5.1.6.3 Institutional Services

Rural credit activities are vital in improving productive resources through purchase of

agricultural inputs, filling the consumption gap whe n it occurs, availing resources for

meeting social obligations, etc. The survey result revealed that 31% of MHH and 17% of

FHH of the sample cases involved in credit market. Both formal and informal institutions are

providing credit activities in the area. The major formal credit providing institution is

Wisdom Micro Finance (WMF), which provides only short-term loans. Out of the

65
beneficiaries of credit service, about 48% of MHH and only 18% of FHH got the service

from the WMF while the rest 52% of MHH and 82% of FHH got credit from informal

sources like friends and local money lenders. This clearly indicates that FHH has less access

to credit service from formal sources.

Even if women bear better credit risks than men (Tiruwork, 1998), since they are unable to

meet collateral requirements such as oxen and land, they are not in a position to take loans.

Women lack access to credit is also constrained by institutional and social biases against

women.

Lack of access to extension services is another constraint that women in the area face. The

adoption of new technologies is often influenced by the farmer’s access to extension

services, since extension agents provide improved inputs and technical advice. In this study,

the frequency of visit by extension agents is strongly associated with the gender of the

farmer. On average, FHH reported fewer visit by extension agents (24.6%) and a larger

proportion of women reported no extension visit at all (75.4%). On the other hand, about

51% of MHH had visit from exte nsion agents during last production season.

It is important to keep in mind that the differences in number of reported visit by extension

agents may not only be attributable to the gender of the farmer but instead could result from

other factors that are correlated with gender of the farmer. For example, it is plausible that

extension agents might prefer to visit farmers with more land or those who have already

adopted improved technology, all of which happened to be correlated with gender.

66
Worldwide extension systems reach more men than women farmers (Saito and Spurling,

1992). Accordingly, studies conducted in India, Indonesia, Philippines and most other

countries showed that most extension agents and the lead farmer they target are men. Male

extens ion agents are generally unaware of the need to communicate differently with women

and view rural women as farmers’ wives not farmers in their own right. Women’s shyness

around men, their lack of access to factors of production and consequently difficulty of

adopting new technology reinforce this misconception. So, male agents tend not to

communicate with women (ibid).

In the study area, there are about 12 extension agents (all male) supervising a total of 8950

households. The extension agents are expected to facilitate fertilizer credit distribution and

collecting repayments in addition to the technical support. Some of them were also involved

in administration activities in their respective PAs.

The result from studies in Africa shows that MHH has more access to yield increasing inputs

(land, labour, fertilizer, improved seed, credit, extension and training) and access to

education and markets (Saito and Surpling, 1992; Ndiyo and Urassa, 2002; Saito et al.,

1994). Others have observed that gender disparities in access to productive resources in

agriculture exist and persist because of legal, social and institutional factors that create

barriers for women (Quisuming, 1995). As discussed above, the results of this study are

consistent with the studies stated above. In general, differential access to and control of

resources between the two sexes in the area seems to be a direct reflection of the culturally

prescribed gender division of labour. For example, as long as milking cows is an exclusive

activity of women, they continue to enjoy full rights to dispose of butter and milk. Likewise,

67
men do ploughing using oxen for production of crops over which they have complete control.

The following table (Table 19) summarizes of the information on access and control of

resources.

Table 19. Access and Control Profile of Resources and Benefits

MHH FHH
Items Male member Female member Male member Female member
Access Control Access Control Access Control Access Control
Land x x x - x x X x
Farm tools x x x - x x X x
Farm oxen x x - - x x X -
Farm inputs x x - - x x X x
Cow x x x x x - X x
Sheep/goat x x x x x x X x
Pack animal x x x - x x X -
Chicken x x xx xx x - xx xx
Livestock x - xx xx x - xx xx
products
Grain x x x x x x X x
Source: Own survey (2003) NB: X =access and control, XX=more access and control

5.1.7 Decision Making in the Household

Women and men in the study area have no equal share regarding decision- making power in

the household and even at the community level. In some cases, the dec ision making power is

closely associated with the pattern of the gender division of labour. For example, farm work

decisions are largely made by men while household maintenance decisions are essentially

made by women. In case where decisions cannot be reached through consensus, the husband

has the final say.

68
Generally speaking, like in most part of the country as stated by Dejene (1994), the opinion

of husband is weigh more than the opinion of the wife in the study area. The bigger the value

of an item, the more the men decide on it. The decision-making power over some of the

important resources and income is briefly discussed and presented in Table 20.

Table 20. Proportion of Households Participating in Decision Making (%)

Activities MHH (N=75) FHH (N=65)

NoΨ Male Female Both χ 2 NoΨ Male Female Both χ2

Sales of livestock 67 38.8 1.5 59.7 39.9* 40 2.5 70 27.5 27.9*

Sale of cash crops 10 80.0 0.0 20.0 3.6** 4 25.0 50 25.0 0.5**

Sale of food crops 71 22.5 4.2 73.2 54.5* 46 0.0 71.7 28.3 8.7*

Use of income 75 37.3 1.3 61.3 41* 64 0.0 73.4 26.6 14.1*

Use of agricultural 67 64.2 0.0 35.8 5.4* 27 3.7 70.4 25.9 18.7*
input
Use of fixed assets 69 31.9 1.4 66.7 44.1* 55 1.8 72.7 25.5 43.1*

Cropping calendar 74 67.6 0.0 32.4 9.1* 60 6.7 63.3 30.0 29.2*

Source: Own survey (2003) NB: *, *=Significant at 1% and 5% probability level, respectively.
Ψ
Number of household involved in the activities listed

Regarding farming operations, husband and wife discuss together when and how to carry out

the different farm operations, though men have the strongest voice in the organization of the

activities such as when to plough, weed, harvest, etc. For example, most of the decisions on

cropping calendar are made by men (67.6%) while the rest 32.4% are made by both men and

women in MHH. In this case, women have no power in deciding on crop calendar in MHH.

However, in FHH about 63.3% of the decision was made by women. Men also have the

highest decision making power on when and how much of the grains should be sold and

69
stored (Table 20).

In the area, men dominate the decision to buy or sell animals such as cow, oxen, horses,

donkey and sheep. Nevertheless, most women said that they are consulted before such a

decision is taken. Sales of livestock products such as eggs and milk are decided by women.

They utilize the income obtained for buying food items such as salt, oil, coffee, etc. Their

proverb goes in line with this, “Sangaafi kan sangaarraa argamu kan dhiiraati, Saaniifi kan

sa’arraa argamu kan dubartiiti”, meaning ox and its products belong to men, as cow and its

products belong to women.

5.2 Econometric Analysis

Under this section, definition of the dependent variable and explanatory variables is briefly

discussed. Subsequently, the hypothesized explanatory variables are analysed with the help

of CD production function and the results interpreted. In addition, MVP of the significant

variables is compared with factor cost to indicate allocative efficiency. Finally, the sources of

agricultural productivity difference between MHH and FHH obtained using Oaxaca

decomposition model are presented.

5.2.1 Definition of Variables and Hypothesis Setting

Dependent variable : The amount of gross value of farm output expressed in Birr per hectare

was used as a dependent variable. Outputs of major crops namely teff, wheat, barely, bean,

maize, sorghum and enset were included in the definition of the dependent variable.

Moreover, income from livestock sold and livestock by-products, if any, in the production

period of year 2002/2003 was included.

70
Explanatory variables: In line with the theoretical background and on the basis of the

previous studies on similar subjects of productivity analysis, the following explanatory

variables were hypothesized to affect the dependent variable.

1. Size of livestock holding: Farmers in the study area consume some percentage of their

livestock and livestock by-products and sell the remaining for cash income generating.

Besides source of cash, livestock increases the availability of manure, which would increase

the productivity of crops. Oxen are the most important source of traction power and farmers

who own oxen would be in a position to undertake farm activities on time. Therefore, it is

expected that a household with a bigger size of livestock, as measured by TLU, will have

opportunity to acquire production inputs and thus improve his/her productivity. Hence a

positive relationship is expected.

2. Land size : Refers to cultivated land expressed in hectare. The size of land reflects

ownership of an important farm asset. Larger farm size implies more resources and greater

capacity to invest in farm and increased production (Ellis, 1988). Thus, farmers with larger

land size could be expected to have higher gross value of farm output.

3. Education of household head: For most of the farming households, the decision on

what/how/how much to produce is made by the household head. Education levels have a

bearing on farmer’s access to improved farm techniques and effective use of information

availa ble on technologies. This implies that the education level of the household head, as

expressed in number of years of schooling, would have positive impact on productivity of

farmers.

71
4. Inorganic fertilizer: The variable stands for all kinds of chemical fertilizer (UREA and

DAP) measured in terms of kg per ha. Chemical fertilizer is used to increase soil fertility and

hence increment in the productivity is expected.

5. Herbicides: This refers to chemicals used in litre per ha to control weeds. It is exp ected to

have positive impact on productivity.

6. Labour: Labour is one of the major inputs for agricultural production. It is expressed in

man-days per hectare. The variable includes all labour spent in the major type of activities on

farm. An increase in man-days increases the production and productivity. Hence, it is

expected to have positive relationship. In this study, labour is categorized as male and female

labour.

5.2.2 Estimation of the Production Function

As discussed in section 4.3.1, CD prod uction function was employed to estimate the

parameters of the production function for MHH, FHH and pooled data set. Before fitting the

data to CD production function, multicollinearity test for explanatory variables was done

using VIF. The results of VIF analysis indicate that the VIF values for all continuous

explanatory variables were by far less than 10. Therefore, all the explanatory variables were

included in the model for further analysis.

72
Table 21. Multicollinearity Test Among Explanatory Variables
Variables Unit R-Square VIF Tolerance
Livestock holding TLU 0.426 1.744 0.574
Land size Ha 0.533 2.140 0.467
Education level Years 0.158 1.188 0.842
Fertilizer use Kg/ha 0.457 1.842 0.543
Herbicide use Litre/ha 0.251 1.336 0.749
Male labour Man-day 0.361 1.565 0.639
Female labour Man-day 0.239 1.314 0.761
Source: Own computation (2003)

The existence of heteroscedasticity problem that violates the assumption of constant variance

was checked. The graph of standardized residuals against the frequency of their occurrence

showed normal distribution of the residuals, which indicate the absence of a

heteroscedasticity problem. Moreover, to test the existence of autocorrelation Durbin

Watson(d) test was applied. The value of Durbin Watson was 2.01. From the Durbin Watson

tables we find that for 140 observations and 7 explanatory variables, the critical values are

d L=1.50 and du=1.75 at 1% (Gujirati, 1995). Since the estimated value of 2.01 lies below

2.25 (4-du) and above 1.75 (d u), the hypothesis that there is no problem of autocorrelation

problem is accepted.

Using equation (3) and (4) discussed in section 4.3.1, the estimates of the CD production

function for male and female headed households are presented in Table 22. All the

production functions were found to be significant as evidenced by significant F-value at 1%

level of probability (Table 22), implying the null hypothesis stating that all the coefficients of

explanatory variables are zero is rejected. The adjusted coefficients of multiple determination

indicate that the variation in gross value of farm output per hectare associated with the

73
factors of production specified in the models was 55.5%, 55.6% and 57.2% in MHH, FHH

and pooled data set, respectively.

In this model, seven explanatory variables were included among which four variables

namely, livestock holding, herbicides use, land size and male labour were statistically

significant for MHH production function while livestock holding, land size, herbicides use

and female labour were significant for FHH production function (Table 22).

Table 22. Parameters of Cobb-Douglas Production Function

Variables Unit Pooled (N=140) MHH (N=75) FHH (N=65)
Coefficient t-value Coefficient t-value Coefficient t-value
Constant 4.654 21.052* 5.271 17.798 * 4.122 11.75*
Livestock holding TLU 0.137 1.871 *** 0.213 2.077* 0.179 1.702***
Land size Ha 0.336 4.381 * 0.383 4.047* 0.306 2.588**
Education level Years 0.052 0.857 0.115 1.385 0.074 0.782
Fertilizer use Kg/ha 0.030 0.401 0.015 0.154 0.003 0.032
* ***
Herbicide use Litre/ha 0.153 2.549 0.156 1.799 0.156 1.725***
Male labour Man-day 0.166 2.386 * 0.352 3.587* 0.124 1.209
*
Female labour Man-day 0.275 4.337 0.073 0.757 0.338 3.264*
Adjusted R2 57.2% 55.5% 55.6%
F-Value 27.52 * 14.19* 12.44*
Durbin Watson 2.01
Source: Model Output (2003) NB: *,** and *** indicate statistical significant at 1%, 5% and 10%
probability level, respectively.

Dependent variable is gross value of farm output per ha and all the variables in the model are

transformed into their logarithmic form. The practical difficulty arises in the conversion of

raw data of value zero to logarithmic form being the logarithm of zero is undefined. To

overcome this problem, the zero observations may be replaced by some figure of arbitrary

74
small size or a constant value can be added to all observations (Heady and Dillon, 1961).

Accordingly, a zero observation was replaced by a small number near to zero (0.001).

It was stated previously in section 4.3.1 that the coefficients of the production function

estimated are called the elasticity coefficient, indicating the percentage share of each

explanatory variable in the variation of the dependent variable or the average percentage

change in the dependent variable as the result of 1% change in one explanatory variable,

keeping other factors constant. Brief discussions of the significant variables in both MHH

and FHH production functions are presented below.

Land size : The size of farmland holding at the disposal of a farmer is one of the most

important variables affecting the level of farmers’ gross value of farm output per ha. It has a

significant and positive impact on productivity of agriculture in both MHH and FHH,

implying that the larger the land size, the higher is the yield. Other factors being constant, a

10% increase in the area under the major crops increases gross value of farm output by

3.83% and 3.06% for male and female headed households, respectively. In other words, these

figures indicate that farmland contributed about 38.3% and 30.6% to the output of the total

inputs for MHH and FHH, respectively.

The possible explanation is that if farmers have more access to land, they are encouraged to

use improved technologies and hence output per ha increases. The correlation coefficient

between land size and use of improved inputs was 0.39, which is statistically significant at

1% probability level.

75
According to a study conducted in Ethiopia by Cropenstedt and Mulat (1997), assuming

other factors remaining constant, more of the variation in productivity would come from the

change in the size of cultivated land. This fact was also supported by the findings of Addis et

al. (2000). They found that land size is the most important factor affecting the productivity

level of agricultur e in East Shoa as indicated by the coefficient of regression 0.56 and 0.76

for MHH and FHH, respectively.

Labour: In this study, the variable that stands for male labour had a significant and positive

impact on the gross value of farm output of MHH. A 10% increase in the amount of male

labour resulted in 3.52% increase in gross value of farm output in MHH, keeping other

factors constant. This clearly shows that an increment of male labour use increases the output

significantly. Moreover, a 10% increase in female labour of the FHH increases output by

3.38%, keeping others factors constant.

The result of this study shows that female labour is much more productive than male labour

in FHH as compared to that in MHH and similarly male labour is more productive in MHH

as compared to FHH. The possible explanations may be female household members in MHH

involved less in agricultural activities as compared to females in FHH since the farmer has

relatively larger male labour. As discussed in section 5.1.2.3, the intensity of male labour use

in MHH was greater than that in FHH and also the intensity of female labour use in

agriculture in FHH was greater than that in MHH. The findings of Saito et al. (1994) in

Kenya and Nigeria, and Udry et al. (1995) in Burkina Faso indicated that labor is an

important factor in determining the output.

76
Herbicide use: Herbicides had positive and significant impact on the productivity of the

MHH and FHH. A 10% increase in herbicides use resulted in 1.56% increase in output in

both MHH and FHH. This shows that an increment in herbicides use increases the output.

This may be because as farmers use more inputs, the productivity per ha of crop increases,

which ultimately increase s the gross value of farm output of the household.

Livestock holding: As discussed earlier, livestock is an integral part of the agricultural

system of the area and thus used as one of the determining factors of output. In this study,

livestock owned as measured by TLU was significant variable for both MHH and FHH. It

has positive impact on output indicating that a 10% increase in TLU has brought about

2.13% and 1.79% increase in output for MHH and FHH respectively, other factors held

constant. The possible explanation is that farmers having more livestock can get more

income from sales of livestock by-products and live animals. Besides, most farmers in the

study area use animal dung to manure enset and other field crops, which could increase the

productivity of the land. In addition, farmers who own oxen could be able to prepare their

land in time and increase yield.

5.2.3 Comparison of Productivity of the two Groups

Identifying the source of the output difference between MHH and FHH is a necessary step in

the determination of an appropriate policy intervention. Many previous studies reviewed in

Quisumbing (1995) have indicated differences in output per acre or per person but have

failed to isolate the source of these differences.

Difference in gender affects the sources of output difference by shifting the value of the scale

and slope parameters of the production functions. In order to test the overall significance of

77
the difference in production function parameters between male and female headed

households and hence the shift in production functions due to gender difference, the Chow’s

F-test was computed using Equation (10) of section 4.3.2.

Table 23. Comparison of the Overall Production Parameters of Both Groups

Item Number of Degree of Residual Sum of Chow’s F-value

observations freedom Squares

MHH 75 67 19.07 2.03 *

FHH 65 57 78.78

Pooled 140 124 110.59

Source: Own computation (2003) NB: *=Significant at 5% probably level

F-value for the overall difference in parameters is significant at 5% level of probability,

implying that there is a shift in production function between the two groups (F

tabulated=1.93). However, in order to test whether the shift is of scale parameter or of slope

parameters, a pooled production function with an intercept dummy for gender of the

household (Equation 6) was estimated (Table 24).

78
Table 24. Estimates of Production Function for Pooled Sample With Dummy Gender
Explanatory Variables Regression coefficients t-value
Intercept 4.660 21.018 *
Livestock holding 0.136 1.845***
Land size 0.344 4.422*
Education level 0.062 0.995
Fertilizer use 0.039 0.508
Herbicides use 0.153 2.543**
Male Labour 0.184 2.457**
Female Labour 0.265 4.053*
Dummy Gender (1 for -0.048 -0.667
MHH, 0=Otherwise)
Adjusted R 2 57%
F-value 24 *
Number of observation 140
Source: Own computation (2003) NB:*,**,***=Significant at 1% , 5% and 10% probability level,

respectively

As shown in Table 24, the coefficient of gender in the pooled production function is

insignificant in the analysis (t=-0.048, p=0.667). This implies that gender yield differential,

apparently, is caused by the difference in the intensity with which measured inputs are

applied in MHH and FHH rather than by differences in the technical efficiency with which

these inputs are used. In other words, FHH in the study area are equally technically efficient

as that of MHH.

Again in order to check whether the shift is due to slope parameters, Chow’s F-value was

computed using pooled function with dummy gender residual. This has resulted in a

significant difference at 5% of probability level in slope coefficients of the production

79
functions of the male and female headed households. Therefore, the differences between the

two groups were due to differences in the slope parameters not due to the intercept.

Table 25. Chow’s Test for Checking the Difference in the Slope Parameters
Item Number of Degree of Residual Sum of Chow’s F-value

observations freedom Squares

MHH 75 67 19.07 1.96 *

FHH 65 57 78.78

Pooled dummy 140 124 110.21

Source: Own survey (2003) NB: *=Significant at 5% probability level

As a concluding remark, the shift in the production function between male and female

headed households was through a shift in the slope, i.e. output elasticity with respect to

various inputs. However, it could not be concluded from these tests whether the slope

coefficients of all the explanatory variables were different or some of them were equal. This

requires the use of dummy variable technique in order to identify which individual slope

parameters are significantly different between the two groups and which parameters are not

(Koutsoyiannis, 1997). Hence, a pooled production function was re-estimated with both the

intercept and slope dummies applying Equation 11.

80
Table 26. Estimates of a Pooled Production Function With Both Intercept and Slope Dummies
Explanatory Variables Regression coefficients t-value
Constant 4.122 15.55*
Livestock holding 0.229 2.253**
Land size 0.400 3.425*
Education level 0.148 1.035
Fertilizer use 0.004 0.042
Herbicides use 0.190 2.283**
Male Labour 0.144 1.600***
Female Labour 0.370 4.320*
Dummy Gender (1=MHH, 0.411 2.053**
0=Otherwise)
Slope dummies
Livestock holding -0.113 -0.674
Land size -0.212 -0.942
Education level -0.088 -0.573
Fertilizer use -0.017 -0.098
Herbicides use -0.079 -0.811
Male labour 0.224 2.054**
Female labour -0.398 -2.316 **
Adjusted R 2 59.7%
F-value 14.71
Sample size 140
Source: Own computation NB: *, **, ***=Significant at 1%, 5% and 10% probability level,

respectively.

The model results show that the slope dummies, with the exception of the dummy variables

for male and female labour, were not statistically significant. Thus, the hypothesis of

homogeneity in the regression coefficients of these inputs between the two production

functions was rejected where as the hypotheses were accepted in the case of all other inputs.

81
5.2.4 Marginal Value Product of Inputs

Allocative efficiency can be determined by comparing the MVP of a factor with its

opportunity cost (factor price). The MVP of a factor is the additional return from adding one

more unit of that factor holding all other inputs constant. MVP, which exceeds its

opportunity cost, suggests that there is scope for productivity raising output by increasing the

use of that factor. Conversely, increasing the use of a factor, which has MVP less than the

associated opportunity cost, decreases the productivity (Ellis, 1988).

The MVP of the inputs used in the production functions was calculated from the elasticity

coefficients (Table 22) and from the geometric mean of input level in the production

processes using Equation (7) in section 4.3.1. For comparison purposes the prices of inputs

were required. Accordingly, the factor price for herbicides was 46 Birr per litre in 2002/2003.

Human labour was valued at their opportunity cost, i.e., the local wage rate of 6 Birr per day.

Local average renting price of a farmland was estimated to be 399 Birr per ha and this value

was used for valuing farmland. The MVP of the significant variables in the CD function is

given in Table 27.

Table 27. MVP and Factor Cost of Significant Explanatory Variables

MHH (N=75) FHH (N=65) Factor cost
Inputs
Unit MVP Ratio MVP Ratio
Land Ha 467 1.17 202 0.51 399
Male labour Man-day 13 2.19 14 2.32 6.0
Female labour Man-das 7 1.22 18 3.11 6.0
Herbicides Litre 327 7.12 147 3.19 46
Source: Own computation (2003)

82
As it can be seen from Table 27, the MVP of farmland and herbicides in MHH were greater

than that of the inputs for FHH while the MVP for female labour, on the other hand, were

greater for FHH than MHH.

The MVP of female labour for FHH is more than two times higher than it is for the MHH.

So, this indicates that the productivity of female labour in FHH is much higher than in MHH.

Utilization of one additional unit of female labour (man-day) resulted in 18 Birr increase in

gross value of output for FHH. It was also shown that the MVP of female labour for both

MHH and FHH was higher than the factor cost, implying efficient use of female labour.

Hence, farmers could increase their productivity by using more female labour for farming.

The MVP of male labour was similar in MHH and FHH and higher than opportunity cost in

both sample groups. This shows that male labour is und er-utilised in both groups. Additional

use of one unit of man-day results in 13 and 14 Birr increase in gross value of output for

MHH and FHH, respectively, keeping other inputs constant. Thus, farmers in the study area

could increase their productivity by using additional male labour. It should be kept in mind

that labour has been defined as man-days for production, not as the total labour force of the

households. Therefore, making better use of the total labour force obliviously can improve

the efficiency.

Regarding farmland, the MVP in MHH was more than twice that of FHH. In FHH, the MVP

was lower than the opportunity cost, implying over utilization of farmland ; however, it was

around one in MHH.

83
Finally, the MVP of herbicides was greater than that of its factor cost. Application of one

additional litre of herbicides could increase the gross value of farm output by 327 Birr for

MHH and 147 Birr for FHH, keeping other factors constant. This indicates that it is possible

to increase the productivity of agriculture by using more units of herbicides in the study area.

5.2.5 Source of Productivity Difference

This section presents estimates of the agricultural productivity differences between male and

female headed households using decomposition model. As discussed in section 4.3.1, this

model will help to measure the percentage contribution of the different effects to agricultural

productivity difference between male and female headed households. This method allows

distinguishing the productivity difference that can be explained by differences in household

endowments and differences in the efficiency of these endowments.

In addition to the estimates of production functions, the decomposition analysis requires the

geometric mean values of different inputs and output. Table 28 presents geometric mean

values of various inputs and output in both MHH and FHH. It is observed that the inputs

used by MHH were higher as compared to FHH for all the explanatory variables used in the

model.

84
Table 28. Geometric Mean of Dependent and Explanatory Variables in the Model
Variables Unit MHH (N=75) FHH (N=65)
Gross farm value Birr/ha 2562.83 1089.50
Livestock holding TLU 3.15 2.19
Land size Ha 2.10 1.66
Education level Years 2.29 1.29
Fertilizer use Kg/ha 67.34 10.60
Herbicides use Lit/ha 1.22 1.16
Male labour Manday 68.61 9.70
Female labour Manday 25.76 19.68
Source: Own computation (2003)

By following the methodology described in the section 4.3.1 (Equation 8), the total sources

of productivity difference were decomposed into output elasticities and input endowments

(Table 29).

Table 29. Decomposition of Productivity Difference between MHH and FHH

Percentage Contribution
Sources of productivity difference Due to output elasticities Due to input endowments

Total estimated difference (68.83%) -23.58 92.41

Livestock holding 2.70 7.73
Land size 3.88 9.10
Education level 1.04 6.60
Fertilizer use -4.33 -2.76
Herbicides use 0.01 0.84
Male labour 51.82 68.92
Female labour -78.69 1.98
Source: Own computation (2003)

From Table 29, it can be seen that the total productivity difference in agriculture between the

two groups was about 68.83%. However, they have different human capital endowment and

85
different access to factors and inputs as discussed in the descriptive part. Inputs use

differentials accounted for 92.41%. This implies that the productivity could be increased by

92.41%, if the FHH could adjust their inputs to the same level of MHH. On the other hand,

the difference in output elasticities was -23.58%. This indicates that productivity difference

as the result of difference in output elasticities is greater for FHH as compared to that of

MHH.

A detailed examination of the contributions made by different inputs in the total productivity

gap between male and female headed households reveals that difference in access to male

labour caused the biggest jump. This further indicates that if FHH could adjust their male

labour to the level of MHH, they can increase their productivity by about 68.92%. Hence,

increasing the access of FHH to male labour could highly increase their productivity in

agriculture in the study area. Descriptive results of this study also show that on average FHH

had only 0.74 men adult while MHH had about 1.54 men adult on average, which was

significant at 1% (t=4.84). Farmland, livestock and education level difference between MHH

and FHH made about 9.1%, 7.73% and 6.6% productivity difference in agriculture,

respectively (Table 29).

Different researchers often argued that women's lack of access to resources results in lower

productivity or inability to respond to economic incentives (Quisumbing, 1996; Gladwin,

1991). However, these results need to be int erpreted with caution, since they do not reveal

how levels of inputs may be raised.

Looking at the contribution made by the output elasticities or change in factor specific

productivity, female labour is more efficient in FHH as indicated by 78.69%, which

86
decreases in productivity gap. On the other hand, male labour is more efficient in MHH

(51.82%) followed by farmland (3.88%) and livestock (2.7%). The productivity differences

as the result of efficiency difference of other inputs such as herbicides use (0.01%) and

education (1.04%) are not significant between MHH and FHH.

87
CHAPTER SIX: SUMMARY AND POLICY IMPLICATIONS

6.1 Summary

The study examined the impact of gender differences on agricultural productivity in Wenchi

district of south west Shoa zone. The data used in this study were collected from 75 MHH

and 65 FHH randomly selected from 6 PAs of the district. Independent t-test was used to test

the differences between MHH and FHH in terms of continuous variables and χ 2-test for

discrete variables. Moreover, CD production function was estimated to measure productivity

difference between MHH and FHH.

With regard to socio -economic characteristics of the sample households, MHH are found to

have relatively larger family size and labour force and more literate household members than

FHH. In addition, land holdings of FHH are much smaller than that of MHH. The result also

shows that MHH used more fertilizer, herbicides and credit compared to the FHH. In general,

women farmers are disadvantaged in access to and control over resources compared to men

farmers. Differences in access and control over resources between men and women seem to

be a direct reflection of culturally prescribed gender division of labor. Legal, social and

institutional factors can also create barriers for women farmers to have access to productive

resources.

Women, in the study area, work for considerably longer hours than men. Clearly, there is

limit to the time and energy that women farmers can apply. Women’s burden hinders their

participation in education, training and/or extension activities. With regard to gender role,

women play a significant role in agricultural production. They participate in all activities

except ploughing using oxen, sowing and enset planting. This shows that agric ultural

88
activities are gender oriented in a certain manner.

Seven variables were hypothesized to affect the gross value of farm output per ha. The results

of the production function analysis revealed that four variables were statistically significant.

Farmland, livestock holding, herbicide use and male labour were statistically significant for

MHH while farmland, livestock holding, herbicide use and female labour were significant for

FHH.

A test of structural break between production functions for MHH and FHH revealed that shift

in production function of MHH was due to difference in coefficients of the production

function. The coefficient of the dummy gender had negative but insignificant effect showing

that there is no difference in the intercept term or no difference in technical efficiency

between MHH and FHH.

The MVP of farmland and herbicide were higher for MHH as compared to FHH while the

MVP of female labour was higher for FHH. Comparison of MVP with the factor cost shows

that MHH could increase productivity by using more herbicides and male labour while FHH

could do so by using more herbicides, male and female labour.

The result of the study further suggests that men’s gross value of output per ha was 68.83%

higher than that of women’s. However, the descriptive statistics of this study show that there

is a difference in household endowments, which have very large overall effects on the

productivity difference between male and female headed households by about 92.41%. If

women had the same human capital and used the same amounts of inputs as men, the value

of their output would increase by some 23.58% over the MHH.

89
Thus, it can be concluded that MHH had higher productivity with the existing input level.

However, FHH would be more productive if they have equal access to inputs as that of

MHH. From the coefficient of gender dummy, it can be concluded that women are equally

technically efficient farm managers as men farmers. However, their productivity is well

below potential. Capturing this potential productivity gain by improving the circumstances of

women farmers would substantially increase food production thereby reducing the level of

food insecurity in the area.

6.2 Policy Implications

Based on the findings of the study, the following points need to be considered as possible

policy implications in order to increase the productivity of farmers in general and that of

women farmers in particular.

1. Increasing output by expanding the area under cultivation is not an option in the study area

because there is little arable land uncultivated. Comparison of MVP with the factor cost

showed that land in the study area is over utilized by both MHH and FHH. Therefore, raising

the productivity of land is of central importance for increasing agricultural productivity

through use of yield increasing inputs such as herbicides. As the estimate of CD production

function indicates, herbicide use significantly affects the productivity of agriculture for

farmers in the study area. Therefore, measures should be taken to ensure timely availability

and use of herbicides in the area.

2. Since livestock holding was one of the significant factors influencing agricultural

productivity, intervention to improve livestock sector should be encouraged through

90
empowering farmers to own livestock through provision of livestock credit. Furthermore,

development of improved livestock feed should be encouraged. From the result of

decomposition model, one could conclude that accessing women farmers with livestock

holding of MHH would increase their productivity by about 7.73%.

3. Women work considerably longer hours than men in the area. Most of their work such as

enset processing requires more energy and consumes time. Thus, technologies that can

reduce the time and energy of women in carrying out these activities should be developed

and disseminated to women. Further research should be conducted on labour saving

production technologies.

91
REFERENCES

Addis Tiruneh. 2000. The Missing Link Between Micro and Macro Level Gend er Dis-
aggregated Economic Data in Economic Policy Formulation and Planning in Ethiopia.
WID/Department of Economics, Unpublished Paper, Addis Ababa University,
Ethiopia.
Addis Tiruneh, Teklu Tesfaye, Hungo Verkuijl and Wilfred Mwangi. 2000. Gender
Differential in Agricultural Productivity Among Smallholders in Ada, Lumu and
Gimichu Weredas of the Central Highlands of Ethiopia. Ethiopian Journal of
Development Research, Vol. 22, No.1, pp. 1-23.
Antle, J. and Capalbo, S. 1988. Agricultural Productivity Meas urement and Explanation:
Resource for the Future. Washington D.C, USA.
Asmare Hagos. 1989. The Impact of Farm Size on the Efficiency of the Farmers’
Cooperatives: The Case of Harar-Zuria Awraja. M.Sc Thesis, Alemaya University,
Ethiopia.
Assefa Mebratu and Mutambirwa, C. 1992. Gender Differences in Time and Energy Costs of
Distance for Regular Domestic Chores in Rural Zimbabwe: A Case Study in the
Chiduku Communal Area. World Development, Vol. 20, No. 11, pp.1675-1683.
Bindlish, V. and Evenson R. 1993. Evaluation of the Performance of T&V Extension in
Kenya. World Bank Technical Paper 208, Africa Technical Department Series,
Washington D.C.
Block, A. 1999. Agriculture and Economic Growth in Ethiopia. Growth Multiplier From a
Four-Sector Simulation Model. Agricultural Economics, Vol. 20, No. 3, pp.241-252.
Bogalech Alemu. 2000 a. Mainstreaming Gender in National Development Programmes.
Economic Focus, Vol. 3, No.3, Ethiopia.
_________. 2000b. Policy Approach to Mainstream Gender in Ethiopia. In: Institutionalizing
Gender Planning in Agricultural Technology Generation and Transfer Process,
pp.126-144, Ethiopia.
Boserup, E. 1970. Women’s Role in Economic Development. Routledge, London.
Buvinic, M. and Gupta, G. 1996.Women-headed Households and Women-maintained
Families: Are They Worth Targeting to Reduce Poverty in Developing Countries?
Economic Development and Cultural Change.

92
Buvinic, M. and Youssef, N. 1978. Women-headed Households: The Ignored Factor in
Development Planning. International Center for Research on Women. Washington
D.C.
Chant, S. 1997. Women- headed Households: Diversity and Dynamics in the Developing
World. Macmillian Press Ltd.
Cheema, A. 1978. Productivity Trend in Manufacturing Industries. Pakistan Development
Review, Vol.17, No. 44, pp.241-267.
CISP (International Committee for the Development of Peoples). 1997. The Female Farmers
Fertile Ground. A Gender Assessment Study of Agricultural Inputs Adoption in Rural
Ethiopia, Ethiopia.
Coelli, T. 1998. An Introduction to Efficiency and Productivity Analysis. Kluwer Academic
Publisher, USA.
Croppenstedt, A. and Mulat Demeke. 1997. An Empirical Study of Cereal Crop Production
and Technical Efficiency of Private Farms in Ethiopia: A Mixed Random Approach.
Applied Econometrics, Vol. 29, pp.1217-1226. In: Sharadaweir Working Paper
Series, University of Oxford.
CSA (Central Statistics Authority). 1998. Statistical Report Vol.1. The 1994 Population and
Housing Census of Ethiopia at Country Level, Ethiopia.
_____________. 2001. Ethiopian Child Labour Survey. Statistical Bulletin, Vol.262,
Ethiopia.
_____________. 2002. Ethiopia Statistical Abstract 2001, Ethiopia.
Davidson, J. 1995. Agriculture, Women and the Land: The African Experience. West View
Special Studies on Africa, West View Press.
Dejene Aredo. 1989. The Gender Division of Labour in Ethiopian Agriculture. A Study of
Time Allocation Among People in Private and Cooperative Farms in Two Villages,
Ethiopia.
____________. 1994. Female -headed Farm Households in two Contrasting Regions in
Ethiopia : Access to and Management of Resources.
Duggan, L. 1997. Households and Family. In: The Women, Gender and Development
Reader, pp. 103-107, Visvanathan, N. (eds), USA.
Elabour, I. 1991.The Impact of Structural Adjustment Program on Women and Their
Households in Bendel and Ogun States, Nigeria. In: Structural Adjustment and

93
 African Women Farmers, pp. 128-150, Gladwin, C.H (eds), USA.
Ellen, K. 1992. Food Security and Child Nutrition: The Interaction of Income and Gender of
Household. World Development, Vol.20, No 8-12.
Ellis F. 1988. Peasant Economics: Farm Households and Agrarian Development. Cambridge
University Press, New York..
Farell, M. 1957. The Measurement of Production Efficiency. J.R Statistics Society, Ser A
120.
Folbre, N. 1991. Women on Their Own: Global Patterns of Female Headship. In: Women and
International Development Annual Report, Vol. 1. Gallin, R. and Ferguson, A. (ed).
Westview Press, USA.
Garfinkel, I. and Mclanahan, S. 1986. Single Mothers and Their Children: A New American
Dilemma. The Urban Institute Press, Washington D.C.
Gittinger, J., Chernick, N., Horenstein, and Saito, K. 1990. Household Food Security and the
Role of Women. World Bank Discussion Paper 96, Washington, D.C.
Grown, C. and Sebstad, J. 1989. Expanding Income Earning Opportunities for Women in
Developing Countries. World Development, Vol. 17, No. 7-12, pp.937-952.
Gujarati, D. 1995. Basic Econometrics, 3rd ed. Mc Graw Hill Inc, New York.
Heady, E. and Dillon J. 1961. Agricultural Production Functions. Iowa State Univers ity
Press, USA.
Ilahi, N. 2000. The Intra- household Allocation of Time and Tasks: What Have we Learnt
from Empirical Literature. Policy Research Report on GAD, Working Paper Series
No. 13.
Jamison, D. and Lau, L. 1982. Farmer Education and Farm Efficiency. Johns Hopkins
University Press for the World Bank.
John, B. and David, D. 1991. Women and Development in the Third World, New York.
Kalirajan, K., Obwona, M and Zhao, S.1996. A Decomposition of Total Factor Productivity
Growth: The Case of Chinese Agricultural Growth Before and After Reform. Amer.
J.Agri.Econ, Vol.78, No.2, pp.331-338.
Koopman, J. 1997. The Hidden Roots for the African Food Problem: Looking Within the
Rural Household. In: The Women, Gender and Development Reader, pp 103-107,
Visvanathan, N. (eds), USA.
Koutsoyiannis, A. 1997. Theory of Econometrics: An Introductory Exposition of

94
 Econometric Methods, 2nd ed. Macmillan Publishers Ltd, USA.
Laura, K. 1991. The Sociology of Gender. St. Martin’s Press, New York, USA
Lisa, G. and Ritu S. 1995. Women’s Local Global Forum. Journal of the Society for
International Development, No. 1-4, p.37.
Martha, F. 1987. Rural Women: Unequal Partners in Development. International Labour
Office, Geneva.
Medina, S., Belen, T. and Eliseo, A. 1994. Filippino Famil ies and Households in Three
Selected Philippine Areas. Philippine Sociological Review.
Mencher, J., and Okongwu, A. 1993. Conclusion: Pulling It All Together. In: Where Did All
the Men Go? Westview Press.
MEDaC (Ministry of Economic Development and Cooperation). 1998. Survey and Current
Economic Conditions in Ethiopia, Unpublished Paper, Addis Ababa.
MoFED (Ministry of Finance and Economic Development). 2002. Sustainable Development
and Poverty Reduction Strategy Paper, Unpublished Paper, Ethiopia.
Moock, P.1976. Managerial Ability in Small Farm Production: An Analysis of Maize Yields
in the Vigha Division of Kenya. Ph.D Dissertation. Columbia University, New York.
NBE.1999. Quarter Bulletins (Various Issues). Addis Ababa, Ethiopia.
Ndiyo, D. and Urassa, K. 2002. Gender Imbalance in Agricultural and Non-agricultural
Activities and Its Impact on Household Food Security: A Case Study of Morogoro
Rural, Tanzania. Sokoine University of Agriculture.
NOP (National Office of Population). 1999. Women in Ethiopia. Addis Ababa, Ethiopia.
Oaxaca, R. 1973. Male-Female Differentials in Urban Labour Markets. International
Economic Review, Vol. 14,No. 1, pp.693-09.
PCICRW (Population Council and International Center for Research on Women). 1988.
Women-headed Households: Issues for Discussion. Paper Prepared for Seminar Series
on the
Determinants and Consequences of Female- headed Households.
Quisumbing, A. 1993. Gender Issues in Agriculture: Overcoming Barriers to Increased
Productivity and Welfare. Population and Human Resource Division, South Asia
Department III. World Bank, Washington D.C, USA.
_________. 1995. Gender Differences in Agricultural Productivity: A Survey of Empirical
Evidence. Food Consumption and Nutrition Division International Food Policy

95
 Research Institute, Washington D.C, USA.
Rosenhouse, S. 1988. Identifying the Poor: Is Headship a Useful Concept? Paper Prepared for
Joint Population Council and International Center for Research on Women.
Saito, K. and Spurling, D. 1992. Developing Agriculture Extension for Women Farmers.
World Bank Discussion Paper, 156. Washington, D.C.
Saito, K., Hailu, Mekonen., and Spurling, D. 1994. Raising the Productivity of Women in
Sub-Saharan Africa. The International Bank for Construction and Development of the
World Bank, Washington D.C, USA.
Seyoum, E., Battese, G. and Fleming, E. 1998. Technical Efficiency and Productivity of
Maize Producers in Eastern Ethiopia: A Study of Farmers Within and Outside the
Sasakawa-Global 2000 Project. Agricultural Economics, Vol. 19, No.3, pp.341-348.
SG. 1999. The Food Chain in Sub-Saharan Africa. Proceedings of the Workshop Held in
Bamako, Mali, October 15-19/1999, Geneva.
Solow, R. 1957. Technical Change and Aggregate Production Function. Review of
Economics and Statistics 24, pp.312-320.
Storck, H., Berihanu Adenew, Bezabih Emana, Borowiecki and Shimelis W/Hawariat. 1991.
Farming System and Farm Management Practices of Smallholders in Hararghe
Highlands. Farming Systems and Resource Economics in the Tropics, Vol.11,
Wirtschafts-verlag Vauk, Kiel, Germany.
Tibaijuka, A. 1994. The Cost of Differential Gender Roles in African Agriculture: A Case of
Smallholder Banana -Coffee Farms in Kangera Region, Tanzania. Agricultural
Economics, Vol. 45, No.1.
Tiruwork Tizazu. 1998. Access to Resources and Productivity of Female- headed Households:
The Case of East Gojjam and North Shoa, MA Thesis, Addis Ababa University,
Ethiopia.
Udry, C. 1994. Gender, Agricultural Productivity and the Theory of the Household.
Discussion Paper. Tinbergen Institute, Amsterdam- Rotterdam.
Udry, C., Aldeman H. and Haddad L. 1995. Gender Differentials in Farm Productivity:
Implications for Household Efficiency and Agricultural Policy. International Food
Policy Research Institute. Washington D.C, USA.
Uma, L. 1991. Women, Structural Adjustment and Transformation. In: Structural Adjustment
and African Women Farmers. Gladwin, C.H (eds), USA, pp.46-80.

96
Visvanathan, N. 1997. Theories of Women, Gender and Development. In: The Women,
Gender and Development Reader. Visvanathan, N., Duggan, L., Nisonoff, L. and
Wiegersma, N. (eds), USA.
World Bank. 2000a. The World Bank Partnerships for Development. Published by Creative
Communicates Group for the World Band, USA.
________. 2000b. Voices of the Poor: Can Any One Hear Us? Deepa Narayan, Raj Patel, Kai
Schafft, Anne Rademacher and Sarah Koch-Schulte (eds). Washington, DC.
_________. 2001. Engendering Development: Through Gender Equality in Rights, Resources
and Voice. Policy Research Report. Washington, DC.
_________. 2003. World Development Indicators. Washington DC
Wudinesh Hailu. 2003. Gender and the Agricultural Economic Sector. In: Gender Policy
Dialogue Series No.1. Gender and Economic Policy, Addis Ababa, Ethiopia.
WVE (World Vision International Ethiopia). 2001. Baseline Survey for the Integrated Rural
Development Intervention, Addis Ababa, Ethiopa.
Zenebework Tadesse. 2000. Revisiting Rural Development through a Gender Lens.
Unpublished Paper ,Addis Ababa University, Ethiopia.

97
 APPENDICES

Appendix 1. Survey Questionaries

I. Background Information
1.1 Name of Enumerator and their signature________ Interview date__________________
1.2 Peasant Association _______________Village/Goti___________________
1.3 Agro ecology ___________1=Dega 2=Woina Dega 3=Kolla

II. Household Characteristics

2.1 Name of the respondent (household head) ___________________________
2.2 Marital Status 1= Married 2= Single 3=Divorced 4=Widow
5=Separated
2.3 If the household head is female, what was the reason? 1=Death of husband
2=Unmarried because of economic problem 3=Husband migration
4=Polygamy 5=Barren 6= Conflict with spouse 7=Others (specify)

2.4 List of family members and their characteristics including the respondent
N Name of Sex Age Relations hip Formal Main Farming Ethnic
o family 1=Male (Years with household Educati occupatio Experienc Group
members 2=Female ) head on n es (Years) (c)
(a) (Years) (b)
1
2

2.5 Religion 1=Orthodox 2=Muslim 3=Protestant 4=Catholic 5=Others

(Specify)
2.6 If no formal education, did you attended basic education and/or religion-based education?
1=Yes 2=No
2.7 If yes, for how many years? _________ (Write ‘0’ for illiterate)

III. Land
3.1 Do you own land? 1=Yes 2=No
3.2 If yes, total land size in senga ____________
3.3 If no, why? ________________________________________
3.4 Indicate your land use pattern
S/N Type Timad owned Timad owned by Timad owned Source
by man woman by both (a)
1 Cultivated
2 Forest and wood land
3 Grazing
4 Waste land
5 Fallowed
6 Others (specify)

a) 1=Husband 2=Wife 3=Daughter 4=Son 5=Others (Specify)

b) 1=Farming 2=Trade 3=Others
c) 1=Oromo 2=Amhara 3=Gurage 4=Others
a
1=Inherited 2=Given by family 3=Allocated by government 4=Rented 5=Others (Specify)

98
3.5 What type of land access do you have?
S/N Type senga Remark
1 Own
2 Rented in
3 Rented out
4 Borrowed
5 Shared in
6 Shared out
7 Inherited
8 Communal
9 Others

3.6 If you rented in land, how much do you pay in this year?_____ (Birr) and for how
long?___ (Years)
3.7 How do you plough your land? __________________ 1=Hand tools 2=Using
own oxen 3=Rented Oxen 4=Borrowed Oxen 6=“Lata” 7=Others
(Specify)
3.8 When did you obtain land (E.C)? ________
3.9 Do you practice soil and water conservation activities (SWC)? 1=Yes 2=No

3.10 If yes, which techniques and area covered?

S/N SWC techniques Area covered (Timad)

1
2
3.11 If no, what is/are the major reasons? (According to their importance)_______________
3.12 Do women have right to obtain land? 1=Yes 2=No
3.13 If no, why? _____________________________________________________________
3.14 Do women have the capacity to direct on the use of land? 1=Yes 2=No
3.15 If no, why? (Explain) ________________________________________________
3.16 Do women inherit land? 1=Yes 2=No
3.17 If no, why? ___________________________________________________________

IV. Crop Production Systems

4.1How many days do you participate in farming in a month on average ?
Male=_______Female=_____

4.2 Please estimate your production level for the last years.
No Crops Cultivated Area (Senga) Production (Qt) in 1995
1995 E.C

99
 4.3Did you apply agricultural inputs during last production period (1995 E.C)? 1=Yes
2=No If yes, fill the following table

Seed (kg) Chemicals used Fertilizer (kg) Labour (Man days)

Crop Improved Local Insecticide Herbicide Pesticides Dap Urea Manure Hired Family
Cultivated (Liter) (Liter) (Gram) Male Female M F

4.4 Which perennial crops do you practice? Coffee=_____

Enset=_______Gesho=______Chat=____

V. Labour Use
5.1 On which agricultural activities do you participate? (Rank according to their importance)
_______________________________________________________________________
5.2 Which agricultural activities are often undertaken by men only in your area? (Rank
according to their importance) __________________________________________
5.3 Which agricultural activities are often undertaken by women only in your a rea? (Rank
according to their importance) __________________________________
5.4 Have you encountered labour shortage during last production years? 1=Yes 2=No
5.5 If yes, for what specific activities have you encountered labour shortage?_________
1=Cultivation of land 2=Weeding 3=Crop harvest 4=Threshing 5=Others
(Specify)
5.6 If yes, how did you overcome problems posed by labour shortage? 1=Hiring labour
2=Assistance from relatives 3=Traditional labour pooling system 4=Was not able to
overcome the problem5=others (specify)
5.7 If you hired labour, how many man-days during the last year? ____________
5.8 What type of labour you hired? Male ___________ Female_____________
5.9 How much do you pay annually if you rent in labour for crop production? ______Birr
5.10 What is an average price of daily labour? _____Birr
5.11 According to your perception who works longer hours both in agriculture and other
tasks? 1=Male ___hours 2=Female____hours

5.12 How long do you and your family spend on average on each of the following activities
during 1995 E.C?
S/ Activities Number of family member participated
N Male Female
I Agriculture
1 Land preparation
2 Ploughing
3 Sowing
4 Weeding
5 Harvesting
6 Threshing
7 Transporting
8 Marketing
9 Vegetable gardening
10 Enset planting
11 Enset processing
12 Livestock herding

100
 S/ Activities Number of family member participated
N Male Female
13 Cleaning of house
14 Milking
II Domestic work
1 Fetching water
2 Grain mill
3 Food preparation
4 Child care
5 Wage work
6 Fire wood collection

VI. Use of oxen

6.1 How many oxen of your own do you use for ploughing? _________
6.2 For which activities did you use oxen? 1=Ploughing 2=Threshing 3=others
6.3 Do you face oxen shortage during your farming practices in last production period? =Yes
2=No
6.4 If yes, how did you cope up with such shortage? 1=Pairing with others
2=Exchanging labour for oxen 3=Using oxen of relatives 4=Hiring oxen
5=Others (specify)
6.5 For how many days did you hired oxen? ____________days
6.6 If you hired oxen, how much did you pay per day? ________Birr, _________ in kind
(specify)

VII. Livestock production

7.1 Do you have livestock? 1=Yes 2=No

7.2 If yes, how many livestock have you owned?

Type Number at Income obtained, if sold any Who owned? 1=Men
present in last year (Birr) 2=Women 3=Family
Oxen
Cow
Young Bulls
Calves
Heifers
Sheep
Goats
Chicken
Horse
Mule
Donkey

7.3 Number of milking cows in 1995 E.C? _______, Milk produced per day______ Liter,
lactation period in month______ month, percentage of milk sold during 1995 E.C______%,

101
7.4 What income have you earned from sale of livestock produces in 1995 E.C?
Type Unit Amount Amo unt sold Unit price (Birr) Remark
Produced
Butter Kg
Eggs No
Hide and skin No
Cheese Kg

7.5 What type of animals are sold by men in your locality? ___________________
7.6 On which livestock do women have the right to sale in your area? (List according to their
importance) ____________________________________________
7.7 Do women have the right to own livestock? 1=Yes 2=No
7.8 If no, why? (Explain the reason) ______________________________________

VIII. Agricultural Extension Services

8.1 Did you participate in extension program in 1995 E.C? 1=Yes 2=No
8.2 Did you have an extension contact? 1=Yes 2=No
8.3 If yes, frequency of contact? (Total number of visit per year)___________
8.4 If no, why? 1=No DA nearby 2=Unknowingly 3=No ne ed for service 4=Others
8.5 If yes, what is/are the types of extension message given by the agents? 1=Use of fertilizer
2=Natural resource advice 3=Animal production advice 4=Use of credit 5=Home
economics 6=Use of insecticides/herbicides 7=Others (specify)
8.6 Are there female development agents? 1=Yes 2=No
8.7 Who do you prefer? 1= Male agent 2=Female agent and why for? _____________
8.8 For how many years have you participated in extension program? _______Years
8.9 How far from your house the DA center? ______Km ________hours
8.10 Have you attended farmer’s training within last years? 1=Yes 2=No
No Training Topics Number of Training provider Remark
days

8.11 Have you ever attended demonstration or any other trials? 1=Yes 2=No
8.12 Have you ever hosted field day or demonstration trial? 1=Yes 2=No
8.13 What are your sources of finance for purchase of agricultural inputs? 1=Crop sales
2=livestock sales 3=Off- farm activities 4=Credit 5=Others
8.14 Do women have access and control over agricultural inputs in the area? 1=Yes
2=No
8.15 If no, why? (Explain) _____________________________________________

IX. Financial Services

9.1 Is there any credit giving institution in the area? 1=Yes 2=No
9.2 If yes, what are they?_____________________________________________
9.3 Did you receive credit services last year? 1=Yes 2=No
9.4 If yes, from where do you get the services? 1=Bank 2=NGO 3=Micro Finance
Institute 4=Local money lender 5=Service cooperatives 6=Friend and relatives
7=Others (specify)
9.5 If no, why? 1=Fear of inability to repay 2=High interest rate 3=Lack of collateral
4=No credit services 5=No need of credit 6=Others (specify)

102
9.6 Why you prefer number 9.4 as source of credit? 1=Low interest rate 2=Accessible
3=low procedure 4=Others (Specify)
9.7 What was the amount you got from credit services during last year? _________Birr
9.8 What is/are collateral provided for credit? 1=Land 2=Future crop 3=Co-signature
4=Building 5=Group guarantee 6=Livestock 7=Others
9.9 For what purposes you have obtained the credit? 1=Purchase of seeds 2=Purchase of
fertilizer/chemicals 3=Purchase of oxen 4=To fill up family requirement 5=to settle
debts 5=Others
9.10 Do other household members get the services? 1=Yes 2=No
9.11 If yes, number of male and female who were getting credit services during last year?
Male=_________Female=_____
9.12 Do you repay the loan on time? 1=Yes 2=No
9.13 If no, why? _______________________________________________________
9.14 Do women have acce ss to credit in the area? 1=Yes 2=No
9.15 If no, why? (Explain the reason) __________________________________________
9.16 Do women have control over the credit they borrowed 1=Yes 2=No
9.17 If yes, to what extent? _____________________________________________
9.18 If no, why? (Explain) ____________________________________________
9.19 What was the repayment period of credit you have taken? 1=Short term
2=Medium term 3=Long term
9.20 What is/are the major problems relating to credit services? ______________________
9.21 Who is getting more credit services in your area? 1=Male 2=Female
9.22 What do you think that the reason for answer of Question 9.21? _______________

X. Sources of Income
10.1 What is your source of income? 1=Sale of crops 2=Sale of livestock 3=Off-farm
income 4=Remittance 5=others
10.2 What was the average price for major crops during the last years?
No Type of crops Average price (Birr) Per Qt Remark

10.3 What is your total food need in a year? _________Qt

10.4 Do you consult your family when you want to spend money? 1=Yes 2=No

XI. Market Accessibility

11.1 Where did you sell your agricultural products? 1= On farm 2=Local markets
3=Others
11.2 To whom do you sale your agricultural products? 1=Wholesaler 2=Retailer
3=Consumers 4=others
11.3 Who usually go to the market from your family? 1=Male 2=Femaleand name of
market____
11.4 How far is the nearest market? _______Km _________Hours
11.5 How far is the longest market? _______Km _________Hours

103
XII. Decision Making
12.1 Who decide on the following in your family? (X)
S/N Activities Men Women Others (Specify)
1 Crop calendar
2 Sale of food crops
3 Use of income
4 Sale of cash crops
5 Use of agricultural inputs
6 Use of improved seeds
7 Sale of livestock
8 Sale of fixed assets
12.2 Do you participate in any leadership position in the PA or district? 1=Yes 2=No
12.3. If yes, explain your responsibility? _________________________________________
12.4 If no, why? ___________________________________________________
12.5 Are you a member of any formal organization other than PAs? 1=Yes 2=No
12.6 If yes, which one? 1=Cooperatives 2=Women’s group 3=Farmer’s group
4=Others
12.7 What services do you get from the organization you belong to? 1=Loan
2=Education 3=Seed 4=Fertilizer 5=Labour 5=Others

12.8 Have you taken gender education? 1=Yes 2=No

12.9 Do women participate in house decisions? 1=Yes 2=No
12.10 Who inherit assets of the household when husband died? 1=Wife 2=Daughters
3=Son 4=others

XIII. Others
13.1 Is there polygamy in the area? 1=Yes 2=No
13.2 What are harmful traditional practices? (Explain each of them if any)
__________________________
13.3 For how many days did you abstain from farming practices in 1995 E.C due to social
engagement and holidays? ____________ Days
13.4 How many days did you spend on communal activities in 1995 E.C? ____Days
13.5 Compare number of female headed before three years and now in your locality?
1=Increasing 2=Decreasing 3=No change
13.6 If it increases, why? _____________________________________________________
13.7 What type of house do you have? 1= Corrugated iron sheet 2= Thatched roof
3=others
13.8 Is there a flourmill in the village? 1=Yes 2=No
13.9 If no, what is the distance from your village to the nearest mill?
Distance in Km ____(Kms) Distance in walking hours _____ (hours)

104
13.10 Access and control over resources in the family member
Resources Access 7 Control8
Male Female Male Female
Land
Farm tools
Farm inputs (Seed, fertilizer,pesticides, etc)
Oxen
Horse/donkey
Cow
Sheep/goat
Chicken
House
Grain
Credit
Livestock products

Appendix 2. Conversion of Livestock Number into TLU

Animal TLU Animals TLU

Chicken 0.013 Young bulls 1.0

Sheep/goat(adult) 0.13 Cows and ox 1.0

Sheep/goat(young) 0.06 Donkey(young) 0.35

Calf 0.25 Donkey(adult) 0.70
Heifers 0.75 Horse 1.1
Source: Source: Storck et al., 1991

Appendix 3. Conversion Factor for Family Member into AE

Age group (years) Male Female

<10 0.6 0.6
10 – 13 0.9 0.8
>13 1 0.75
Source: Storck et al., 1991

7
1=More 2=Less 3=No

105
Appendix 4. Conversion Factor for household Labour into ME

Age group (years) Male Female

<10 0.0 0.0
10 – 13 0.2 0.2
14-16 0.5 0.4
17-50 1 0.8
>50 0.7 0.5
Source: Storck et al., 1991

106
 Declaration

I, the undersigned, declare that this thesis is my work and all reviewed materials used for the

thesis have been duly acknowledged.

Name : Wakweya Tamiru Yada

Signature : ________________

Place : Alemaya University

Date of Submission: June 2004

107